Highway 406 Roundabout Positive Guidance Analysis

Transcription

1 . Highway 406 Roundabout Positive Guidance Analysis DRAFT REPORT: November 12, 2010

2 DRAFT REPORT: Highway 406 Roundabout Positive Guidance Analysis TABLE OF CONTENTS 1 INTRODUCTION THE SPECIFIC GOAL OF THE PROJECT THE 2- PART PROCESS GENERAL PART 1: REFINEMENT OF THE POSITIVE GUIDANCE DESIGN TO BE TESTED PART 2: TESTING OF THE PROPOSED POSITIVE GUIDANCE PLAN RESULTS OF THE LABORATORY SIMULATION SUGGESTED CHANGES TO MTO PLAN...11 APPENDIX A: RECOMMENDED CHANGES TO POSITIVE GUIDANCE PLAN...1 APPENDIX B: LABORATORY SIMULATION REPORT...1 LIST OF FIGURES FIGURE 1: WESTBOUND APPROACH POSITIVE GUIDANCE PLAN (EAST OF TUNNEL ENTRANCE)...4 FIGURE 2: WESTBOUND APPROACH POSITIVE GUIDANCE PLAN (WEST OF TUNNEL)...5 FIGURE 3: SOUTHBOUND APPROACH POSITIVE GUIDANCE PLAN...6 FIGURE 4: EASTBOUND APPROACH POSITIVE GUIDANCE PLAN...7 ii

3 DRAFT REPORT: Highway 406 Roundabout Positive Guidance Analysis 1 INTRODUCTION The work documented in this report was carried out in fulfillment of one of the key risk management recommendations from a report entitled Highway 406/East Main Street Intersection Options: Human Factors and Road Safety Review prepared by Delphi MRC and Human Factors North. The design of a modern roundabout at the intersection of Highway 406 / East Main Street is difficult in part due to the placement of the roundabout on a crest. The situation is further complicated by the fact that the westbound approach to the junction passes through a tunnel under the Welland Canal, resulting in a steep grade on that section of roadway as it approaches the roundabout. Depending on the final design of the roundabout, these challenges on the westbound approach may have a significant influence on the need for, and ability of, westbound drivers arriving from the east to select the appropriate lane for their entry into the multi-lane roundabout. Based on these concerns, a more detailed and specific exploration from a human factors standpoint using driver simulation was recommended. Delphi-MRC was engaged to carry out this project utilizing the services of the University of Guelph s Driving Simulator and Human Factors North. The essence of the project was to use the driving simulator and a representative sample of drivers to help refine the design of key elements of the positive guidance plan for the approaches to a roundabout located at the terminus of Highway 406. This report must be read in conjunction with the detailed Laboratory Simulation Report, which constitutes Appendix B. 1

4 DRAFT REPORT: Highway 406 Roundabout Positive Guidance Analysis 2 THE SPECIFIC GOAL OF THE PROJECT The simulation process took place in three distinct steps: 1. In an initial process, a proposed positive guidance plan was developed by MTO and Ourston Roundabouts Canada. This plan was reviewed in detail by IBI and MTO staff using the driving simulator to help visualize specific changes that led to a refined plan that was submitted for the more detailed driver behaviour testing. 2. A carefully structured and selected group of drivers representing younger, middle-aged, and older drivers drove the simulated routes approaching the roundabout. 3. Both observational analyses and a post-drive questionnaire were used to develop an understanding of driver behaviour on the approaches to the roundabout. The goal of this project and the detailed evaluation of driver behaviour and performance in the simulator were to assess the effectiveness of the proposed positive guidance plan for the roundabout design under consideration and to use information gleaned from the analysis to identify necessary improvements to the proposed positive guidance plan. 2

5 DRAFT REPORT: Highway 406 Roundabout Positive Guidance Analysis 3 THE 2- PART PROCESS 3.1 General As noted earlier, the development of the final positive guidance plan took place in two parts: an initial discussion and analysis phase involving IBI and MTO staff and intended to develop the base positive guidance plan to be used in testing drivers; and the actual driver testing. Each of these phases of the work is discussed separately below. 3.2 Part 1: Refinement of the positive guidance design to be tested In refining the positive guidance plan to be tested, the approaches to the roundabout were computer modeled, based on drawings and other documents, and then driven by IBI and MTO staff members to determine the effectiveness of a range of sign and pavement marking treatments. During this design visualization and interaction phase, a number of design decisions were made based on these interactions with the model of the Highway 406 roundabout used in the driving simulator. These decisions included: Replacing sign images on the upstream gantry approaching the tunnel from westbound Highway 27 with those provided by MTO. Relocating the overhead exit sign for the roundabout bypass ramp on westbound Hwy 27 (west of the tunnel) to the start of the painted gore. Relocating the roundabout warning sign, yield ahead sign, and advance roadway identification signs on the southbound Highway 406 approach 130 m upstream. Please note that implementing this change caused these signs to interfere with the blue tourism signs and the 60 km/h speed zone signs. As a result, these signs were also moved upstream 130 m. Shortening the simulation run-out times after traversing the roundabout or bypassing the ramp. Developing an additional scenario to assess the movement from northbound Wellington Street through the roundabout to northbound Highway 406. Details on the final positive guidance plan and roundabout approach configurations carried forward to the testing phase of this assignment are outlined in the following figures. 3

6 DRAFT REPORT: Highway 406 Roundabout Positive Guidance Analysis Figure 1: Westbound approach positive guidance plan (east of tunnel entrance) 4

7 DRAFT REPORT: Highway 406 Roundabout Positive Guidance Analysis Figure 2: Westbound approach positive guidance plan (west of tunnel) 5

8 DRAFT REPORT: Highway 406 Roundabout Positive Guidance Analysis Figure 3: Southbound approach positive guidance plan 6

9 DRAFT REPORT: Highway 406 Roundabout Positive Guidance Analysis Figure 4: Eastbound approach positive guidance plan 7

10 DRAFT REPORT: Highway 406 Roundabout Positive Guidance Analysis 3.3 Part 2: Testing of the proposed positive guidance plan Once the design was finalized it was evaluated by having 48 participants, stratified into the age groups of 18 to 24, 25 to 54, and 55 and older, drive the roundabout approaches in all directions. Participants initially drove a 3 to 5 minute practice drive in the simulator. Drivers were then required to find their way to various locations in the Welland/Highway 406 region using each of the seven segments listed below: Drive 1a: Westbound drivers on Regional Road 27 heading to Highway 406 North Drive 1b: Westbound drivers on Regional Road 27 heading to Welland Drive 2a: Northbound drivers on Farr Road turning left and heading to Highway 406 North Drive 2b: Northbound drivers on Farr Road turning left and heading to Welland Drive 3: Eastbound drivers on East Main Street heading to Highway 406 North Drive 4: Eastbound drivers entering East Main Street from Wellington Street heading to Highway 406 North Drive 5: Southbound drivers on Highway 406 heading to Port Colborne The drives were selected to help determine whether the proposed positive guidance results in the following driver behaviours: Drive 1a: Westbound drivers on Regional Road 27 heading to Highway 406 North moving into the right lane well before reaching the tunnel Drive 1b: Westbound drivers on Regional Road 27 heading to Welland moving into the left lane well before reaching the tunnel Drive 2a: Northbound drivers on Farr Road turning left and heading to Highway 406 North moving promptly into the rightmost lane before reaching the tunnel Drive 2b: Northbound drivers on Farr Road turning left and heading to Welland moving promptly into the leftmost lane before reaching the tunnel Drive 3: Eastbound drivers on East Main Street heading to Highway 406 North moving into the left lane on the approach to the roundabout and turning left through the roundabout Drive 4: Eastbound drivers entering East Main Street from Wellington Street heading to Highway 406 North moving into the left lane on the approach to the roundabout and turning left through the roundabout Drive 5: Southbound drivers on Highway 406 heading to Port Colborne taking either lane on entry to the roundabout (both lanes will take drivers to Port Colborne) 8

11 DRAFT REPORT: Highway 406 Roundabout Positive Guidance Analysis In total, the drive was approximately 21 minutes long and was driven in seven separate segments. A number of simulator measures were collected during participants drives, including: velocity, location of lane change manoeuvres, lane position on the approaches to and within the roundabout, and the effects of practice. Finally, participants were interviewed at length about how well they understood the design elements and whether they had any difficulties with certain signs. 9

12 DRAFT REPORT: Highway 406 Roundabout Positive Guidance Analysis 4 RESULTS OF THE LABORATORY SIMULATION 1 The positive guidance on the westbound approach up to the ramp bypass following the tunnel appears to have worked well. A ruse to force drivers initially into the wrong lane for their destination, so that they would have to make a visible choice to move to the correct lane, was only partially successful, because there were reasons other than a desire to reach the destination which could have lead drivers to move back into the correct lane (i.e. preference to stay in right lane, selection of left lane after turning left due to accepted driver behaviour). The ruse was most successful on Drive 1b where drivers were forced into the right lane. Here the majority of drivers initiated their lane change to the left lane at 750 m from the tunnel, suggesting they were using the sign information. On this drive, most drivers were in the correct lane by 600 m from the tunnel, and all by 350 m from the tunnel. The positive guidance plan was also successful in leading almost all drivers (92 out of 96 drives or 96%) heading to Hwy 406 N to take the ramp bypass; only one driver (on each of 2 drives) went into the roundabout and exited there. Another 2 entered the roundabout but due to forgetting their destinations, incorrectly exited to Welland. The positive guidance plan on the approaches to the roundabout itself appears to have been somewhat less successful. On the westbound approach, all drivers exited the tunnel in the correct lane (left) for their destination. By the time they entered the roundabout however, on Drive 1b 20% of them, and on Drive 2b 18%, had moved to the right lane, an unnecessary lane change, suggesting uncertainty about the correct lane. Although the vast majority of drivers entered the roundabout in a lane which accessed their destination, this may well have been due to the fact that on this three leg roundabout, drivers could reach their destinations on both southbound and both westbound entry lanes. When there were two lanes, only one of which was correct for the destination, as on the eastbound approach, drivers chose the wrong lane about one quarter to one third of the time. On the eastbound approach the two guide signs, both with complex messages, were separated only by 30 m (3 sec at travel speed), which likely gave drivers insufficient time to digest the information. On the westbound approach the last two guide signs on the roundabout approach were separated by 80 m (5 sec at travel speed). This short separation may not have been a problem because drivers could access their destination by means of either lane. Driver comments indicate that a number were confused by the lane markings in the roundabout. Within the roundabout, depending on the drive, 25 to 68% (southbound) of drivers made one or more unnecessary lane changes, that is lane changes not required to get into the correct lane, suggesting a high degree of uncertainty about the appropriate lane, especially on the southbound approach. On the southbound approach the two map signs were only separated by 4 seconds at travel speed. The signs were complex and created a high load in a fairly short time period. 1 This section is based almost exclusively on the contents of the Executive Summary of the full Laboratory Simulation Report provided in Appendix B. 10

13 DRAFT REPORT: Highway 406 Roundabout Positive Guidance Analysis 5 SUGGESTED CHANGES TO MTO PLAN Based on the findings from this analysis, the following modifications to the current MTO positive guidance plan are recommended: Reverse the order of the two map signs on the southbound approach, from higher level (Welland) to lower level (RR27). In the Vscope test, subjects selecting a lane for a given destination on this approach timed out on 2.2 to 28.9% of trials. The usual approach to wayfinding is to identify higher level destinations first, e.g. a city, and then roads within that city. Doing so by reversing the order of the map signs may make these signs easier to comprehend reducing reading time requirements. Provide greater separation of the two southbound guide signs by pulling the first sign about 4 sec (approximately 100 m) further from the roundabout. As a result of moving this sign, the upstream guide and warning signs proposed on this approach will also need to be relocated. An analysis of the speeds driven on the southbound approach in the simulator indicates drivers are likely to have only 4 seconds after passing the first map sign to deal with the second one. This is very limited time for a complex sign. While subjects were able to make decisions within 2 to 3 seconds in the Vscope test, they were not driving at the time. Also many of them timed out, taking more than 8 seconds to make decisions especially on roundabout-related signs. Greater separation of these two signs, and reduction of content, would allow drivers more time to read them. Remove "City Centre/Centre Ville" from the southbound Welland map sign and the roadway identification turn-off sign (mounted on splitter island). Remove the ground mounted lane designation signs (white on black) on all approaches. A test of which signs were remembered best after the simulation indicated that the ground-mounted, regulatory lane designation signs were not well remembered. Subjects also rated their confidence that they had seen these signs in the simulation as low. Both findings suggest that drivers did not allocate much attention to these signs. They are visible at the same time as the overhead lane designation sign and therefore compete for the subject s attention. Because they are ground mounted and to the sides of the roadway, it is not as clear which lane is referred to as is the case for the overhead lane designation signs. Provide greater separation of the ground mounted map and the overhead lane designation signs on the eastbound approach, by moving the ground mounted sign to the west by about 1 sec of travel. Currently, drivers only have 3 sec to deal with the overhead signs after passing the ground mounted map sign. Although this located the map sign closer to the Wellington Street intersection, traffic volumes turning from Wellington Street are relatively low and many of those drivers will be stopped before turning and will have a chance to look at the sign while stopped. Remove the reference to Regional Road 27 from the map sign to Welland on the westbound approach. On this approach, all drivers exited the tunnel in the correct lane (left) for their destination. But by the time they entered the roundabout however, about one in five had moved to the right lane - an unnecessary lane change. This suggests uncertainty about the correct lane. The ground mounted map sign in combination with the overhead lane designation signs had the highest rate of 11

14 DRAFT REPORT: Highway 406 Roundabout Positive Guidance Analysis timeouts in the Vscope test (44%). If the roundabout were a standard intersection, Welland would be signed, but not Regional Road 27. Route markers could be used before and after the roundabout. Re-design the lane markings within the roundabout. Driver comments indicate that a number were confused by the lane markings in the roundabout. Before the three lanes actually start, there are two sets of white dots that go half way between the new lanes and the western entrance to the roundabout. Subjects commented that it made it difficult to tell exactly where the lanes were going. Within the roundabout many drivers made unnecessary lane changes, this is, lane changes not required to get into the correct lane. The percentage of drivers making unnecessary lane changes varied from a low of 25% on Drive 4 - Eastbound from Welland to Highway 406 North, to a high of 58% on Drive 5 Southbound from highway 406 to Regional Road 27. On Drive 5, 34% of drivers made two or more unnecessary lane changes, suggesting a high degree of uncertainty about the appropriate lane. Further analysis of the Dutch style roundabout lane designation arrows should be considered. For the roundabout signs with poor comprehension scores, a common error was to assume that they simply meant that there would be a curve in the road ahead. In their comments, a few drivers commented that the presence of the dot in the centre clarified the meaning to some extent, but others indicated that they did not understand the significance of the dot. Dutch (hook-style) roundabout signs were posted in the lab from a previous iteration of the study, and some of these were used as distracter items in the immediate recognition test. It should be noted that the hook-style signs give a stronger impression of the tight radius to be encountered than the signs used in the study. Some drivers commented that the Dutch-style signs were easier to understand than the ones used in the study. These recommended modifications are summarized on a plan appended to this report as Appendix A. It must be emphasized that the simulation tested a very specific roundabout design with specific signing and marking. The outcomes in terms of lane changing and lane selection may be very different were this configuration and its signing and marking to be changed. 12

15 DRAFT REPORT: Highway 406 Roundabout Positive Guidance Analysis APPENDIX A: RECOMMENDED CHANGES TO POSITIVE GUIDANCE PLAN A - 1

16 DRAFT REPORT: Highway 406 Roundabout Positive Guidance Analysis Figure A-1: Westbound approach recommended positive guidance plan A - 2

17 DRAFT REPORT: Highway 406 Roundabout Positive Guidance Analysis Figure A-2: Southbound approach recommended positive guidance plan A - 3

18 DRAFT REPORT: Highway 406 Roundabout Positive Guidance Analysis Figure A-3: Eastbound approach recommended positive guidance plan A - 4

19 DRAFT REPORT: Highway 406 Roundabout Positive Guidance Analysis APPENDIX B: LABORATORY SIMULATION REPORT B - 1

20 FINAL REPORT ON HIGHWAY 406 SIMULATION Alison Smiley, Ph.D., CCPE Tom Smahel, M.Arch., CCPE Lana Trick, Ph.D. Ryan Toxopeus, M.Sc. November 12, 2010 UNIVERSITY OF GUELPH

21 EXECUTIVE SUMMARY INTRODUCTION The goal of this study was to evaluate the effectiveness of the positive guidance to be provided for drivers through a proposed multilane roundabout termination of Highway 406 South at Regional Road 27 (East Main Street) in the Welland area. The roundabout is at a particularly challenging location, in that the westbound approach is through a tunnel under the Welland Canal with limited sight distance to the roundabout area. The southbound approach is a transition zone from a posted 100 km/h freeway. The University of Guelph Driving Simulator was used to visualize the design of the roundabout and its approaches in order to test the effect of the positive guidance plan on a sample of 48 drivers, young, middle-aged and old, most of whom had minimal, if any, prior experience with roundabouts. DISCUSSION Positive Guidance Findings The simulation was organized into a series of seven drives, intended to help determine whether the positive guidance proposed by MTO results in: 1. Westbound drivers on Regional Road 27 heading to Hwy 406 N moving into the right lane well before reaching the tunnel (Drive 1a) 2. Westbound drivers on Regional Road 27 heading to Welland moving into the left lane well before reaching the tunnel (Drive 1b) 3. Northbound drivers on Farr Road turning left and heading to Hwy 406 N moving promptly into the rightmost lane before reaching the tunnel (Drive 2a) 4. Northbound drivers on Farr Road turning left and heading to Welland moving promptly into the leftmost lane before reaching the tunnel (Drive 2b) 5. Eastbound drivers on East Main Street heading to Hwy 406 N moving into the left lane on the approach to the roundabout and turning right as directed within the roundabout (Drive 3) 6. Eastbound drivers entering East Main Street from Wellington Street heading to Hwy 406 N moving into the left lane on the approach to the roundabout and turning right as directed within the roundabout (Drive 4) 7. Southbound drivers on Hwy 406 heading to Regional Road 27 taking either lane on entry to the roundabout (both lanes will take drivers to Regional Road 27) (Drive 5) Following the simulated drives, subjects completed a sign comprehension questionnaire, carried out a V Scope reaction time test to selected signs and provided general feedback on sign effectiveness. Westbound & North to Westbound Approaches (Drives 1a, 1b, 2a, 2b) The first question was whether the positive guidance plan was effective in getting drivers heading to Hwy 406N into the right lane well before entering the tunnel. This was tested in Drives 1a (Regional Road 27 to Highway 406 North) and 2a (Farr Road NB to Hwy 406N). On the northbound to westbound approach to the roundabout (Drives 1a and 1b), drivers encountered two ground mounted signs indicating a left turn for the destinations of Welland and Hwy 406 to St. Catharines. Northbound drivers who turned left to go westbound, as well as UNIVERSITY OF GUELPH Page ii

22 westbound drivers on Regional Road 27 encountered an overhead lane designation sign indicating Hwy406 N to St. Catharines in the right lane and Welland in the left lane 145 m to the west of the Farr Road intersection. A second overhead lane designation sign was located on the tunnel entrance, 800 m from Farr Road. Figure ES-1 shows sign layouts, legibility distance of each sign in the simulator (where visual resolution is less than the eye in real life), measured speed in the simulator and legibility distance from point of first legibility to the decision point, at the measured speed. Figure ES-1: Guide signs on westbound approach There were a total of 96 trials for drivers heading to Highway 406 North. Each of 48 drivers approached the tunnel twice, once from Regional Road 27 westbound (Drive 1a), and once from Farr Road northbound. (Drive 2a). A stalled vehicle was used to force drivers initially into the wrong lane (left lane) for their destination, so that they would have to change lanes. On Drive 1a, 2/3 of drivers changed lanes back to the right lane immediately after passing the stalled vehicle in Zone 1 before the Farr Road intersection. The guide sign placed 145 m west of the Farr Road intersection could only be read at 104 m away, about 40 m after crossing the Farr Road intersection. Thus most drivers started their lane changes before they could actually read the guide sign either due to a desire to return to the right lane, or as a result of practice on previous drives. On Drive 2a, drivers encountered a stalled vehicle in the left turn lane. All of the drivers went to the right lane, with most lane changes being initiated immediately after the Farr Road intersection, just before most drivers could likely read the guide sign. Thus although most drivers were in the correct lane when entering the tunnel, it seems to have been because they preferred to be in the right lane. UNIVERSITY OF GUELPH Page iii

23 In all but 1 out of 96 trials drivers were fully in the correct lane at least 450 m before the tunnel. The remaining driver drove in the eastbound lanes after turning onto Regional Road 27, into the tunnel, finally returning to the correct side of the road just before the roundabout. In summary, the positive guidance on the westbound approach up to the ramp bypass following the tunnel appears to have worked well. A second positive guidance question was whether drivers would use the bypass lane instead of the roundabout. Mean speed within the tunnel was 83 km/h for both Drives 1a and 2a. With respect to accessing the bypass lane to Highway 406 North, drivers slowed by about 16 km/h (nearly 20%) as they exited the tunnel. This slowing may have been in response to the appearance of the ramp decision point located approximately 50 m beyond the tunnel, equivalent to just over 2 seconds at the speed drivers were travelling. On 92 out of 96 trials, drivers took the bypass ramp; on 2 out of 96 trials the same driver went on to the roundabout and took the exit to Hwy 406 N there. Two trials involved errors in that drivers took the roundabout exit to Welland, not the correct destination. This was due to forgetting in both cases. In summary, the positive guidance plan on the westbound approach was also successful in leading almost all drivers heading to Hwy 406 N to take the ramp bypass. A third positive guidance question on the westbound and north to westbound approaches was the effectiveness of the positive guidance plan in getting drivers heading to Welland into the left lane well before entering the tunnel. This was tested in Drives 1b (Regional Road 27 WB to Welland and 2b (Farr Road NB to Welland). There were a total of 96 trials for drivers heading to Welland. Again a stalled vehicle was used to force drivers initially into the wrong lane (the right lane) for their destination, so that they would have to change lanes. In Drive 1b the majority of drivers initiated their lane change to the left lane at 750 m from the tunnel, suggesting they were using the sign information. In Drive 2b all drivers initiated the change to the left lane in the intersection. Since this was the lane that drivers are supposed to turn into, and since drivers could not have read the sign from this point, it does not provide any comment on the quality of the sign information. In all but 1 out of 96 trials drivers were fully in the correct lane at least 350m before the tunnel. The remaining driver completed the lane change inside the tunnel. The driver who did this had driven through the tunnel correctly on two previous occasions. Thus the plan was successful in leading almost all westbound drivers to Welland to choose the correct lane into the tunnel. The positive guidance plan on the approaches to the roundabout itself appears to have been somewhat less successful. All drivers exited the tunnel in the correct lane (left) for their destination. But by the time they entered the roundabout however, about one in five had moved to the right lane, an unnecessary lane change, suggesting uncertainty about the correct lane. On Drive 1b, drivers slowed steadily on exiting the tunnel, from 84 km/h in the tunnel to 20 km/h (slowest) at 25 m from the roundabout entry. On Drive 2b, behaviour was similar with speeds being reduced from 80 km/h to 20 km/h. Interestingly, on both Drives 1b and 2b, drivers also slowed after the tunnel, when the ramp bypass lane and sign became visible, by 6 9 km/h, even though this was not their destination. The amount of slowing was less than the slowing evident on Drives 1a and 2a (16 km/h), for which the ramp bypass was the destination. With respect to behaviour while in the roundabout, on Drive 1b and on Drive 2b, about one third made an unnecessary lane change. With the current three leg roundabout configuration, either entry lane allowed drivers to reach their destination. UNIVERSITY OF GUELPH Page iv

24 The unnecessary lane changes both on the approach to the roundabout and while in the roundabout suggest uncertainty about which lane was correct. As it happens both lanes were correct, nonetheless unnecessary lane changing is undesirable. Eastbound Approaches to Hwy 406 N (Drives 3 and 4) A fourth positive guidance question was the effectiveness of the positive guidance plan in getting eastbound drivers heading to Highway 406 N into the left lane and taking the second exit to the right. On the eastbound approach to the roundabout (Drive 3), drivers encountered a ground mounted map sign followed very shortly by an overhead guide sign. See Figure ES-2. Of 48 drivers heading for Hwy 406N, 37 entered correctly in lane 1 or 2, but a substantial number (23%) entered the roundabout incorrectly in lane 3. A second eastbound approach involved initially going northbound on Wellington St., turning right (eastbound) onto Regional Road 27 and then taking the roundabout to Hwy 406 N (Drive 4). This drive allowed less lead time, as compared to Drive 3, to determine which was the correct lane to enter the roundabout. Again a substantial number of drivers entered incorrectly (31%) in lane 3. While in the roundabout on Drive 3, 40%, and on Drive 4, 29%, made an unnecessary lane change. These unnecessary lane changes reflect uncertainly about which lane to be in. These findings suggest problems with the positive guidance plan for eastbound drivers. UNIVERSITY OF GUELPH Page v

25 Figure ES-2: Guide signs encountered on eastbound approach Southbound Approach (Drive 5) A fifth positive guidance question was the effectiveness of the positive guidance plan in getting southbound drivers heading to Port Colborne into either entry lane and taking the second exit to the right. On the southbound approach to the roundabout (Drive 5), drivers were given a destination of Regional Road 27 eastbound. Drivers encountered two ground mounted map signs followed very shortly by an overhead lane designation sign. There is a two lane approach with a third added as a slip lane to Welland in the last 135 metres (see Figure ES-3). Figure ES-3: Guide signs on southbound approach With the current three leg roundabout, either entry lane allowed drivers to reach their destination. Of 48 drivers, 3 (6%) forgot their destination and incorrectly took the slip lane to Welland and one driver took the roundabout backwards. UNIVERSITY OF GUELPH Page vi

26 Driver comments indicate that a number were confused by the lane markings in the roundabout. Before the three lanes actually start, there are two sets of white dots that go half way between the new lanes and the western entrance to the roundabout. Subjects commented that it made it difficult to tell exactly where the lanes were going. Within the roundabout 68% of southbound drivers made unnecessary lane changes, and 34% made two or more unnecessary lane changes, suggesting a high degree of uncertainty about the appropriate lane. A few drivers went around the roundabout more than once; a few took the bypass ramp when it was not their destination; even fewer entered in the wrong direction. The combination of a very high information load in the 400 metres on the roundabout approach due to 3 complex guide signs, two ground-mounted and one overhead, providing two destination names for each of three destinations, together with an unusual lane configuration at the entry, likely also contributed to driver confusion. In previous studies it has been found that when drivers are challenged with too much information, they actually take in less information than they would had the information load been lower to start with (Smiley, Donderi, & Smahel, 2005). The amount of lane changing in the roundabout is of concern, suggesting problems with positive guidance. Overall Findings Drivers entering the roundabout are required to yield to the traffic in the roundabout. The positions of the two other vehicles in the roundabout varied randomly from trial to trial and driver to driver. If a participant driver was involved in a collision with a simulated driver within 10 meters of the yield line, a collision was counted against them. These collisions were infrequent, one each on Drives 1b and 2b, and two on Drive 3. Collisions at the entrance to the roundabout were assessed because these were collisions in which the participant driver was definitely at fault. Although collisions occurred in other sections of the roundabout, they are not reported because the simulated drivers did not react appropriately to the participant drivers who drove more slowly than they did. This problem occurred because DriveSafety (designers the tile for the roundabout) did not have an adequate model of driver behaviour in roundabouts to use to control the behaviour of simulated drivers. Factors in Interpretation of the Results In interpreting these results, it must be remembered that first, participants were selected to have minimal exposure to roundabouts: 90% of the sample reported taking roundabouts less frequently than once every 6 months and 31% said that they had never driven in a roundabout. Second, signs in the simulator were legible at shorter distances than is the case in the real world, increasing information load. For example, the map signs on Hwy 406S had letter heights of 350 mm, which would make them legible to the 85 th percentile driver at 168 m.; in the simulator they were legible at m. Third speeds are likely to be lower than in the real world. This would increase the information rate and reduce the impact of poorer visual resolution in the simulator. Fourth, minimal traffic was present, thus drivers who were uncertain about what to do had fewer cues from other traffic than is generally the case. Fifth, in two cases drivers going to Hwy 406N forgot their assigned destination, and in three cases drivers going to Regional Road 27 did the same and took the wrong exit. This is likely an artefact of the test procedure which involved 7 drives in the area, with changing destinations, something which is unlikely to happen in real life within such a short time period. It should be noted that simulators provide useful information concerning relative speeds to be expected in the real world, but not on absolute speeds. For example, at 2 km from the roundabout entry on the southbound approach, the mean speed was 95 km/h. Based on actual speed measurements on a number of 400 series freeways posted at 100 km/h, mean speeds at this point are more likely to be 110 km/h, about 15 km/h higher. UNIVERSITY OF GUELPH Page vii

27 Drivers carried out parts of each of the three approaches more than once. Not surprisingly, there was evidence of a small amount of improvement with practice, though the effects were weak and most did not achieve statistical significance. One exception was Drive 1b, where drivers changed lanes significantly earlier as a result of previous experience with Drive 1a (both drives involved navigating Regional Road 27 east of Farr Road). However there were no practice effects in terms of driving speeds in the approach leading up to the roundabout or ramp, for Drive 1b. Post Simulator Sign Testing Sign Comprehension Questionnaire Signs were assessed as to how confident drivers were that they had seen the sign in the driving simulation ( distractor as well as target signs were used), how easy they were to read, to understand, and how well they prepared drivers for what was coming up ahead. Conventional guide signs performed best and roundabout regulatory signs, worst. An analysis of variance revealed that on average, roundabout-related signs were significantly harder to read and understand, and were judged as significantly worse preparation than conventional guide signs (p <.001). Post-hoc tests revealed the same pattern emerged when comparing each of the individual roundabout signs against the average for the conventional signs in terms of ease of reading and understanding. A test of which signs were remembered best indicated that the ground-mounted, regulatory lane designation signs were not well remembered. Subjects also rated their confidence that they had seen these signs in the simulation as low. Both findings suggest that drivers did not allocate much attention to these signs. They are visible at the same time as the overhead lane designation sign and therefore compete for the subject s attention. Because they are ground mounted and to the sides of the roadway, it is not as clear which lane is referred to as is the case for the overhead lane designation signs. V Scope Reaction Time Test Overall, the best performing signs with respect to average response time and the percent of timeout trials were conventional overhead guide signs on the westbound approach. Overhead roundabout lane designation signs performed better than ground-mounted map guide signs. With the one exception of drivers taking the slip lane to Welland to the right, response times to right destinations were faster than those for left or through destinations. The reduced time for right destinations (the first choice point) may reflect the greater wayfinding effort involved in using a roundabout to reach through and left destinations especially for unfamiliar drivers. The average response time to determine lane choice was significantly longer when there were two correct alternatives as compared to one (p <.001). In addition, there was a significantly higher proportion of timeout trials when there were two correct alternatives as compared to when there was only one (26.7% vs. 11.1%, p <.001). Feedback on Sign Usability Overall When drivers were asked about the guide signs after the simulator runs were complete, only 23% said that there were exactly the right number of signs ; 29% said that there were sometimes too many signs; 19%, sometimes too few, and 29% said sometimes too many and sometimes two few. The area with too many signs was most frequently identified as on the highway immediately before the roundabout, and the area with too few was most frequently UNIVERSITY OF GUELPH Page viii

28 identified as on the highway a kilometre before the roundabout. These questions were not specific to a particular approach. CONCLUSIONS The positive guidance on the westbound approach up to the ramp bypass following the tunnel appears to have worked well. A ruse to force drivers initially into the wrong lane for their destination, so that they would have to make an active choice to move to the correct lane, was only partially successful, because there were reasons other than a desire to reach the destination which could have lead drivers to move back into the correct lane (i.e., preference to stay in right lane, selection of left lane after turning left due to accepted driver behaviour). The ruse was most successful on Drive 1b where drivers were forced into the right lane. Here the majority of drivers initiated their lane change to the left lane at 750 m from the tunnel, suggesting they were using the sign information. The positive guidance plan was successful in leading almost all drivers (92 out of 96 drives or 96%) heading to Hwy 406 N to take the ramp bypass and almost all drivers (95 out of 96 drives) heading to Welland to select the correct lane. The positive guidance plan on the approaches to the roundabout itself appears to have been somewhat less successful. While all drivers exited the tunnel in the correct lane (left) for their destination. By the time they entered the roundabout however, about one in five had moved to the right lane, an unnecessary lane change, suggesting uncertainty about the correct lane. For westbound drivers the last two guide signs on the roundabout approach were separated by 80 m (5 sec at travel speed). This short separation may have contributed to the unnecessary lane changes. Drivers still reached their destination because it was possible to do so by means of either lane. On the eastbound approach only one of two entry lanes was correct for the destination. Here drivers chose the wrong lane about one quarter to one third of the time. On the eastbound approach the two guide signs, both with complex messages, were separated only by 30 m (3 sec at travel speed), which likely gave drivers insufficient time to digest the information. On the southbound approach both entry lanes were correct for the destinations given. However while in the roundabout 68% of drivers made an unnecessary lane change, and 34% of drivers made two or more unnecessary lane changes, suggesting a high degree of uncertainty about the appropriate lane. On the southbound approach the two map signs were only separated by 4 seconds at travel speed. The signs were complex and created a high load in a fairly short time period. Driver comments indicate that a number were confused by the lane markings in the roundabout. Thus information overload and confusion about the lane markings may have contributed to the unnecessary lane changes. It must be emphasized that the simulation tested a very specific roundabout design with specific signing and marking. The outcomes in terms of lane changing and lane selection may be very different were this configuration and its signing and marking to be changed. UNIVERSITY OF GUELPH Page ix

29 RECOMMENDATIONS Based on the simulator drives and on the post-simulator testing, the following recommendations are made. 1. Reversal in order of the two ground-mounted map signs on the southbound approach, from higher level (Welland) to lower level (RR27) In the V Scope test, subjects selecting a lane for a given destination on this approach timed out on 2.2 to 28.9% of trials. The usual approach to wayfinding is to identify higher level destinations first, e.g. a city, and then roads within that city. Doing so, by reversing the order of the map signs, may make these signs easier to comprehend reducing reading and decision time requirements. 2. Greater separation of the two ground-mounted map signs on the southbound approach by moving the first sign about 4 seconds of travel distance or 100 m further north of the roundabout 3. Removal of "city centre/centre ville" from the Welland sign An analysis of the speeds driven on the southbound approach in the simulator indicates drivers are likely to have only 4 seconds after passing the first map sign to deal with the second one. This is very limited time for a complex sign. While subjects were able to make decisions within 2 to 3 seconds in the V Scope test, they were not driving at the time. Also many of them timed out, taking more than 4 seconds to make decisions especially on roundabout-related signs. Greater separation of these two signs, and reduction of content, would allow drivers more time to read them, and should help reduce unnecessary lane changing. 4. Removal of the ground mounted lane designation signs on southbound, westbound and eastbound approaches A test of which signs were remembered best after the simulation indicated that the groundmounted, regulatory lane designation signs were not well remembered. Subjects also rated their confidence that they had seen these signs in the simulation as low. Both findings suggest that drivers did not allocate much attention to these signs. They are visible at the same time as the overhead lane designation sign and therefore compete for the subject s attention. Because they are ground mounted and to the sides of the roadway, it is not as clear which lane is referred to as is the case for the overhead lane designation signs. 5. Greater separation of the ground mounted map and the overhead lane designation signs on the eastbound approach, by moving the first to the west by the equivalent of 1 second of travel. With the current location, the driver has only 3 seconds to deal with the second sign set after passing the first. The V Scope test showed a high rate of timeouts on the eastbound approach. Greater separation of these two signs, would allow drivers more time to read them. While this will result in the first map sign being closer to Wellington St., the bulk of the traffic into the roundabout will be on East Main St. Many of the Wellington St. drivers will be stopped before turning and will have a chance to look at the sign while stopped UNIVERSITY OF GUELPH Page x

30 6. Remove RR27 from the map signs to Welland on the westbound approach On this approach, all drivers exited the tunnel in the correct lane (left) for their destination. But by the time they entered the roundabout however, about one in five had moved to the right lane, an unnecessary lane change, suggesting uncertainty about the correct lane. The groundmounted map sign in combination with the overhead lane designation signs had the highest rate of timeouts in the V Scope test (44%). If the roundabout were a standard intersection, Welland would be signed, but not RR27. Route markers could be used before and after the roundabout 7. Redesign of lane markings within the roundabout Driver comments indicate that a number were confused by the lane markings in the roundabout. Before the three lanes actually start, there are two sets of white dots that go half way between the new lanes and the western entrance to the roundabout. Subjects commented that it made it difficult to tell exactly where the lanes were going. Within the roundabout many drivers made unnecessary lane changes, that is, lane changes not required to get into the correct lane. The percent of drivers making unnecessary lane changes varied from a low of 25% on Drive 4 eastbound from Welland to Hwy 406N, to a high of 58% on Drive 5 southbound from Hwy406 to Regional Road 27. On Drive 5, 34% of drivers made two or more unnecessary lane changes, suggesting a high degree of uncertainty about the appropriate lane. 8. Dutch-style roundabout lane designation arrows should be tested in a future study. For the roundabout signs with poor comprehension scores, a common error was to assume that they simply meant that there would be a curve in the road ahead. In their comments, a few drivers commented that the presence of the dot in the centre clarified the meaning to some extent, but others indicated that they did not understand the significance of the dot. Dutch (hook-style) roundabout signs were posted in the lab from a previous iteration of the study, and some of these were used as distractor items in the immediate recognition test described earlier. It should be noted that the hook-style signs give a stronger impression of the tight radius to be encountered than the signs used in the study. Some of the drivers commented that the Dutchstyle signs were easier to understand than the ones used in the study. UNIVERSITY OF GUELPH Page xi

31 TABLE OF CONTENTS 1 INTRODUCTION METHODS POSITIVE GUIDANCE PLAN DRIVE DESCRIPTIONS EXPERIMENTAL DESIGN MEASUREMENT ZONES MEASURES SUBJECTS DRIVING SIMULATOR RESULTS RAMP DRIVE 1A: WEST ON RR 27, RAMP TO HWY 406 NORTH Lane Change Velocity Wayfinding Errors RAMP DRIVE 2A: NORTH ON FARR ROAD, WEST ON RR 27, RAMP TO HWY 406 NORTH Lane Change Velocity Profile Bypass Ramp Decision Wayfinding Errors ROUNDABOUT DRIVE 1B: WEST ON RR 27 TO WELLAND Lane change Velocity profile Collisions in Roundabout Lane Use in Roundabout Wayfinding Errors ROUNDABOUT DRIVE 2B: NORTH ON FARR ROAD, WEST ON RR 27 TO WELLAND Lane Change Velocity Profile Collisions in Roundabout Lane Use in Roundabout Wayfinding Errors ROUNDABOUT DRIVE 3: EAST ON MAIN ST TO HWY 406 NORTH Velocity Profile Collisions in Roundabout Lane Choice in Roundabout Wayfinding Errors ROUNDABOUT DRIVE 4: NORTH ON WELLINGTON, EAST ON MAIN ST TO HWY 406 NORTH Velocity Profile Collisions in Roundabout Lane Choice in Roundabout Wayfinding Errors ROUNDABOUT DRIVE 5: SOUTH ON HWY 406, TAKE ROUNDABOUT EAST TO RR Velocity Profile Collisions in Roundabout Lane Choice in Roundabout Wayfinding Errors PRACTICE EFFECTS ACROSS SIMILAR DRIVES Location of the Lane Change in Drives 1a and 1b Speed Reductions on Roundabout Approach and Number of Stops DRIVER SIGN QUESTIONNAIRE IMMEDIATE RECOGNITION TEST SIGN QUESTIONNAIRE...48 UNIVERSITY OF GUELPH Page xii

32 4.3 V SCOPE REACTION TIME TESTING Westbound Approach Southbound Approach Eastbound Approach DRIVER FEEDBACK General Comments on Number and Location of Signs Driver Comments on Signs DISCUSSION POSITIVE GUIDANCE FINDINGS Westbound & North to Westbound Approaches (Drives 1a, 1b, 2a, 2b) Eastbound Approaches to Hwy 406 N (Drives 3 and 4) Southbound Approach (Drive 5) OVERALL FINDINGS FACTORS IN INTERPRETATION OF THE RESULTS POST SIMULATOR SIGN TESTING Sign Comprehension Questionnaire V Scope Reaction Time Test Feedback on Sign Usability Overall CONCLUSIONS RECOMMENDATIONS REFERENCES...74 UNIVERSITY OF GUELPH Page xiii

33 LIST OF TABLES TABLE 1: DRIVER DEMOGRAPHICS...10 TABLE 2: DRIVER FAMILIARITY WITH WELLAND...11 TABLE 3: DRIVER EXPERIENCE WITH ROUNDABOUTS...11 TABLE 4: GUIDE SIGNS ENCOUNTERED ON DRIVE 1A...12 TABLE 5: ZONE IN WHICH DRIVERS ADOPT APPROPRIATE LANE...13 TABLE 6: LOCATION OF LANE CHANGES RELATIVE TO TUNNEL ENTRANCE...14 TABLE 7: VELOCITY PROFILE BY ZONE...16 TABLE 8: GUIDE SIGNS ENCOUNTERED ON RAMP DRIVE 2A...17 TABLE 9: LOCATION OF LANE CHANGES RELATIVE TO TUNNEL ENTRANCE...18 TABLE 10: VELOCITY PROFILE BY ZONE...19 TABLE 11: GUIDE SIGNS ENCOUNTERED ON ROUNDABOUT DRIVE 1B...21 TABLE 12: ZONE IN WHICH DRIVERS ADOPT APPROPRIATE LANE...22 TABLE 13: LOCATION OF LANE CHANGES RELATIVE TO TUNNEL ENTRANCE...23 TABLE 14: VELOCITY PROFILE ON APPROACH TO ROUNDABOUT...24 TABLE 15: DRIVER BEHAVIOUR IN THE ROUNDABOUT...25 TABLE 16: GUIDE SIGNS ENCOUNTERED ON ROUNDABOUT DRIVE 2B...27 TABLE 17: ZONE IN WHICH DRIVERS ADOPT APPROPRIATE LANE...28 TABLE 18: LOCATION OF LANE CHANGES RELATIVE TO TUNNEL ENTRANCE...29 TABLE 19: VELOCITY PROFILE ON APPROACH TO ROUNDABOUT...30 TABLE 20: DRIVER BEHAVIOUR IN THE ROUNDABOUT...31 TABLE 21: GUIDE SIGNS ENCOUNTERED ON ROUNDABOUT DRIVE TABLE 22: VELOCITY PROFILE ON APPROACH TO ROUNDABOUT...34 TABLE 23: WHAT DRIVERS DID IN THE ROUNDABOUT...36 TABLE 24: GUIDE SIGNS ENCOUNTERED ON ROUNDABOUT DRIVE TABLE 25: VELOCITY PROFILE ON APPROACH TO ROUNDABOUT...38 TABLE 26: WHAT DID THE DRIVERS DO?...40 TABLE 27: GUIDE SIGNS ENCOUNTERED ON ROUNDABOUT DRIVE TABLE 28: VELOCITY PROFILE ON APPROACH TO ROUNDABOUT...42 TABLE 29: DRIVER BEHAVIOUR IN THE ROUNDABOUT...44 TABLE 30: COMPARISON BETWEEN SIGN TYPES...49 TABLE 31: ROUNDABOUT GUIDANCE CATEGORY ALL SIGNS...50 TABLE 32: ROUNDABOUT REGULATORY CATEGORY ALL SIGNS...51 TABLE 33: STANDARD GUIDANCE CATEGORY ALL SIGNS...52 TABLE 34: FACING NORTHBOUND DRIVERS, GUIDE SIGN ON FARR ROAD...55 TABLE 35: FACING WESTBOUND DRIVERS, FIRST GUIDE SIGN WEST OF FARR RD...56 TABLE 36: FACING WESTBOUND DRIVERS, SECOND GUIDE SIGN WEST OF FARR RD...57 TABLE 37: FACING WESTBOUND DRIVERS, MAP SIGN ON APPROACH TO ROUNDABOUT...58 TABLE 38: FACING WESTBOUND DRIVERS, LANE DESIGNATION SIGN ON APPROACH TO ROUNDABOUT...59 TABLE 39: FACING SOUTHBOUND DRIVERS, FIRST MAP SIGN ON APPROACH TO ROUNDABOUT...60 TABLE 40: FACING SOUTHBOUND DRIVERS, SECOND MAP SIGN ON APPROACH TO ROUNDABOUT...61 TABLE 41: FACING SOUTHBOUND DRIVERS, LANE DESIGNATION SIGN ON APPROACH TO ROUNDABOUT...62 TABLE 42: FACING EASTBOUND DRIVERS ON APPROACH TO ROUNDABOUT...63 UNIVERSITY OF GUELPH Page xiv

34 LIST OF FIGURES FIGURE 1: MTO POSITIVE GUIDANCE PLAN EAST OF TUNNEL...2 FIGURE 2: MTO POSITIVE GUIDANCE PLAN FOR WESTBOUND APPROACH TO ROUNDABOUT...3 FIGURE 3: MTO POSITIVE GUIDANCE PLAN FOR SOUTHBOUND APPROACH TO ROUNDABOUT...4 FIGURE 4: MTO POSITIVE GUIDANCE PLAN FOR EASTBOUND APPROACH TO ROUNDABOUT...5 FIGURE 5: MEASUREMENT ZONES...7 FIGURE 6: ROUNDABOUT GEOMETRY...8 FIGURE 7: DRIVE 1A INCLUDING SPEED MEASUREMENT ZONES...12 FIGURE 8: LOCATION OF LANE CHANGES RELATIVE TO TUNNEL ENTRANCE...14 FIGURE 9: VELOCITY PROFILE BY ZONE...15 FIGURE 10: RAMP DRIVE 2A INCLUDING SPEED MEASUREMENT ZONES...16 FIGURE 11: LOCATION OF LANE CHANGES RELATIVE TO TUNNEL ENTRANCE...18 FIGURE 12: VELOCITY PROFILE BY ZONE...19 FIGURE 13: ROUNDABOUT DRIVE 1B, INCLUDING SPEED MEASUREMENT ZONES...20 FIGURE 14: LOCATION OF LANE CHANGES RELATIVE TO TUNNEL ENTRANCE...22 FIGURE 15: VELOCITY PROFILE ON APPROACH TO ROUNDABOUT...23 FIGURE 16: DRIVER LANE CHOICE IN ROUNDABOUT...25 FIGURE 17: ROUNDABOUT DRIVE 2B INCLUDING SPEED MEASUREMENT ZONES...26 FIGURE 18: LOCATION OF LANE CHANGES RELATIVE TO TUNNEL ENTRANCE...28 FIGURE 19: VELOCITY PROFILE ON APPROACH TO ROUNDABOUT...29 FIGURE 20: DRIVER LANE CHOICE IN ROUNDABOUT...31 FIGURE 21: ROUNDABOUT DRIVE 3 INCLUDING SPEED MEASUREMENT ZONES...32 FIGURE 22: VELOCITY PROFILE ON APPROACH TO ROUNDABOUT...34 FIGURE 23: DRIVER LANE CHOICE IN ROUNDABOUT...35 FIGURE 24: ROUNDABOUT DRIVE 4, INCLUDING SPEED MEASUREMENT ZONES...36 FIGURE 25: VELOCITY PROFILE ON APPROACH TO ROUNDABOUT...38 FIGURE 26: DRIVER LANE CHOICE IN ROUNDABOUT...39 FIGURE 27: ROUNDABOUT DRIVE 5, INCLUDING SPEED MEASUREMENT ZONES...40 FIGURE 28: VELOCITY PROFILE ON APPROACH TO ROUNDABOUT...42 FIGURE 29: DRIVER LANE CHOICE IN ROUNDABOUT...43 FIGURE 30: LOCATION OF LANE CHANGE MANOEUVRES DRIVE 1A AND 1B...45 FIGURE 31: LOCATION OF LANE CHANGE MANOEUVRES DRIVE 2A AND 2B...46 FIGURE 32: PRACTICE EFFECTS ON VELOCITY DRIVE 1B AND 2B...47 FIGURE 33: PRACTICE EFFECTS ON VELOCITY DRIVE 3 AND FIGURE 34: PROPORTION OF TIMEOUT TRIALS AS A FUNCTION OF NUMBER OF CORRECT ALTERNATIVES...54 FIGURE 35: GUIDE SIGNS ON WESTBOUND APPROACH...66 FIGURE 36: GUIDE SIGNS ENCOUNTERED ON EASTBOUND APPROACH...69 FIGURE 37: GUIDE SIGNS ON SOUTHBOUND APPROACH...69 UNIVERSITY OF GUELPH Page xv

35 1 INTRODUCTION The goal of this study was to evaluate the effectiveness of the positive guidance to be provided for drivers through a proposed multilane roundabout termination of Highway 406 South at Regional Road 27 (East Main Street) in the Welland area. The roundabout is at a particularly challenging location, in that the westbound approach is through a tunnel under the Welland Canal with limited sight distance to the roundabout area. The southbound approach is a transition zone from a posted 100 km/h freeway. The University of Guelph Driving Simulator was used to visualize the design of the roundabout and its approaches in order to test the effect of the positive guidance plan on a sample of 48 drivers, young, middle-aged and old, most of whom had minimal if any prior experience with roundabouts. 2 METHODS 2.1 Positive Guidance Plan A positive guidance plan to assist drivers in negotiating the roundabout area was provided by MTO. An overview of the proposed signing plan and pavement marking details are shown in the Figures 1 to 4. UNIVERSITY OF GUELPH Page 1

36 Figure 1: MTO positive guidance plan east of tunnel UNIVERSITY OF GUELPH Page 2

37 Figure 2: MTO positive guidance plan for westbound approach to roundabout UNIVERSITY OF GUELPH Page 3

38 Figure 3: MTO positive guidance plan for southbound approach to roundabout UNIVERSITY OF GUELPH Page 4

39 Figure 4: MTO positive guidance plan for eastbound approach to roundabout UNIVERSITY OF GUELPH Page 5

40 Other positive guidance features incorporated into the simulation include: Supplemental pavement markings Arrows on Farr Road left turn lanes Solid centre line through tunnel Traffic signal at Farr Road intersection with Regional Road 27 Landscaping features in the roundabout centre island Tall elements (fence line and shrubs) in the roundabout have been shown to induce earlier slowing on the approach 2.2 Drive Descriptions Drivers were required to find their way to various locations in the Welland/Highway 406 region. Throughout this report, the seven drives will be referred to as follows: Drive 1a: Westbound drivers on Regional Road 27 heading to Highway 406 North Drive 1b: Westbound drivers on Regional Road 27 heading to Welland Drive 2a: Northbound drivers on Farr Road turning left and heading to Highway 406 North Drive 2b: Northbound drivers on Farr Road turning left and heading to Welland Drive 3: Eastbound drivers on East Main Street heading to Highway 406 North Drive 4: Eastbound drivers entering East Main Street from Wellington Street heading to Highway 406 North Drive 5: Southbound drivers on Highway 406 heading to Port Colborne The drive was approximately 21 minutes long and consisted of 2 minutes filler followed by 1 minute per scenario for a total of (1 + 2) minutes x 7 runs = 21 minutes. The drive was shown in 7 separate segments. In addition there was a practice drive of 3 to 5 minutes. The practice drive did not include a roundabout, but included curves, grades and a left turn to identify drivers who are likely to get simulator sickness. The drives were selected to help determine whether the proposed positive guidance results in: Drive 1a: Westbound drivers on Regional Road 27 heading to Highway 406 North moving into the right lane well before reaching the tunnel Drive 1b: Westbound drivers on Regional Road 27 heading to Welland moving into the left lane well before reaching the tunnel Drive 2a: Northbound drivers on Farr Road turning left and heading to Highway 406 North moving promptly into the rightmost lane before reaching the tunnel Drive 2b: Northbound drivers on Farr Road turning left and heading to Welland moving promptly into the leftmost lane before reaching the tunnel UNIVERSITY OF GUELPH Page 6

41 Drive 3: Eastbound drivers on East Main Street heading to Highway 406 North moving into the left lane on the approach to the roundabout and turning left through the roundabout Drive 4: Eastbound drivers entering East Main Street from Wellington Street heading to Highway 406 North moving into the left lane on the approach to the roundabout and turning left through the roundabout Drive 5: Southbound drivers on Highway 406 heading to Port Colborne taking either lane on entry to the roundabout (both lanes will take drivers to Port Colborne) 2.3 Experimental Design One positive guidance plan was tested. All subjects did all seven drives. The order of drives was randomized, counterbalanced by age and gender. 2.4 Measurement Zones The drives involve traversing a series of zones (see Figure 5). Selected measures of driver behaviour were made in each zone. Figure 5: Measurement Zones A description of the length and location of each zone is given below: Zone 1: 200 metres before the intersection of Farr Road and Regional Road 27. For Drives 1a and 1b, this 200 metres is on Regional Road 27. For Drives 2a and 2b, this 200 metres falls on Farr Road. Zone 2: metres before the tunnel to the tunnel under the Welland canal to the tunnel (Drives 1a, 2a, 1b, 2b) Zone 3: metres inside the tunnel (Drives 1a, 2a, 1b, 2b) Zone 4: 47.5 metres after the tunnel but before the ramp (Drives 1a, 2a, 1b, 2b) UNIVERSITY OF GUELPH Page 7

42 Zone 5: 200 metres travelling east bound on East Main Street towards the roundabout (Drive 3, Drive 4 is travelling that direction but misses Zone 5 because Wellington St. turns onto East Main Street shortly before the roundabout) Zone 6: 200 metres travelling southbound on Highway 406 towards the roundabout (Drive 5) For drives that involve roundabouts, there is also a description of the 200 metres immediately before the roundabout and also inside the roundabout. The roundabout is divided into lanes (1 = centre, 2 = further right, 3 = yet further right) travelling through 4 checkpoints (1 = east, 2 = south, 3 = west, 4 = north) (see Figure 6). Figure 6: Roundabout geometry 2.5 Measures The measures for each zone are enumerated below: Westbound Movements: Zone 1 East of Farr Road on RR 27; South of RR 27 on Farr Road Proportion of drivers who turn right at Farr Road intersection (expected to be 0) Zone 2 Farr Road to Tunnel Entrance Proportion of drivers who change lanes in response to first guide sign west of Farr Road Proportion of drivers who change lanes in response to second guide sign west of Farr Road UNIVERSITY OF GUELPH Page 8

43 Location (metres from tunnel) at which lane change is completed relative to tunnel entrance Zone 3 Inside Tunnel Number of illegal lane changes Zone 4 Tunnel Exit to Roundabout Proportion of drivers who make lane changes after tunnel Location at which lane change is completed relative to bypass ramp Proportion of drivers who take bypass ramp vs. roundabout/intersection for drivers whose destination is Hwy 406 Speed and speed variance at 25 m spacing from tunnel exit to roundabout yield line (profile to indicate distance in m. at which signs are legible, signs are passed, roundabout is visible, 406 on ramp is passed, relative to yield line at roundabout) Eastbound Movements Zone 5 Eastbound approach on Main Street to roundabout Proportion of drivers in the right lane and in the centre lane at various distances at 25 m spacing from the roundabout entrance up to the earliest point at which the relevant guide sign can be read Proportion of drivers who stop at roundabout when they have the right of way Proportion of drivers who turn left incorrectly Southbound Movements Zone 6 Southbound approach on Hwy 406 to roundabout Mean and standard deviation of speed at 2 km, 1km, 200 m before the roundabout and then every 25 m until the roundabout Proportion of drivers who turn left incorrectly Effect of practice Behaviour change with respect to lane selection with subsequent scenarios i.e., is there an order effect (1a vs. 1b; 2 a vs. 2b prior to tunnel (irrespective of destination) Post test questionnaire Understanding of access to Hwy 406 via ramp and at roundabout/ confusion re: bypass Understanding of which lane to be in for Welland vs. Hwy 406 Understanding of right of way in roundabout Understanding of left turn after passing island in roundabout Any of the guide signs unclear (reference static image of each sign) Any of the lane markings unclear (reference static images) At the start of each drive, drivers were forced into the wrong lane by means of a stalled vehicle so that a lane change was required to access the destination they have been assigned. In scenarios 1 and 2, drivers approached the Farr Road intersection on a green light, to minimize the time that they had to read signs and make decisions. UNIVERSITY OF GUELPH Page 9

44 2.6 Subjects A total of 48 subjects in 3 age groups participated in the simulation. Sixteen subjects in each of 3 age groups (18-24, 25-54, 55 years and up) provided experimental data. Of the 48 subjects, 26 were male and 22 were female; 44 held full licenses and 4 held G2 licenses. A total of 66 subjects were tested, but 17 developed simulator sickness. Although the standard predictors were used to screen for simulator sickness, the screen is not perfect, and there were some people of the people who did not report the predictors of simulator sickness nonetheless got sick. The high rates of simulator sickness were to be expected given the large number of turns and the number of seniors included in the study (a high risk group for simulator sickness). One participant deliberately avoided taking roundabouts even if it meant going to the wrong location; data from this participant was dropped. Thus, the data from 18 participants had to be dropped, leaving 48 participants for the study. Driver demographic data is shown below in Table 1. The average driver age was 41 years and the average distance driven per day was 47 km. Table 1: Driver demographics Question N Minimum Maximum Average SD (Standard deviation) Age in years Distance/day driver (km) Familiarity with roundabout rules of the road (1 minimum, 9 maximum) Drivers were asked how often they had been to Welland; nearly half of all drivers (47.9%) had never been to Welland (see Table 2). UNIVERSITY OF GUELPH Page 10

45 Table 2: Driver familiarity with Welland Frequency Percent Never been to Welland Been to Welland 1-5 times Been to Welland 6+ times Total Drivers were asked how often they had driven through a roundabout. Nearly one-third (31.6%) had never used a roundabout (see Table 3). Table 3: Driver experience with roundabouts Frequency Percent Use roundabout once a week Use roundabout once a month Use roundabout once every 6 months Use roundabout once a year Use roundabout less than once a year Never used a roundabout Total DRIVING SIMULATOR RESULTS 3.1 Ramp Drive 1a: West on RR 27, Ramp to Hwy 406 North An illustration of Drive 1a is shown below in Figure 7 and the guide signs encountered on this drive are shown in Table 4. UNIVERSITY OF GUELPH Page 11

46 Figure 7: Drive 1a including speed measurement zones As shown in Table 4, the first sign west of Farr Road was located 613 m from the tunnel entrance and was legible at a distance of 104 m; at a speed of 70 km/h (recorded in the simulator) this sign was legible 36.9 seconds prior to reaching the decision point. The second sign west of Farr Road was located at the tunnel entrance and was legible at a distance of 101 m; at a speed of 70 km/h this sign was legible 5.2 seconds prior to reaching the decision point. The first sign west of the canal was located at the gore of the Highway 406 northbound bypass and was legible at a distance of 120 m; at a speed of 66 km/h this sign was legible 6.5 seconds prior to reaching the decision point. Table 4: Guide signs encountered on Drive 1a Ramp Drive 1a Distance from Decision Point (m) Legibility Distance (m) Speed (km/h) Seconds Legible From Decision Point (sec) UNIVERSITY OF GUELPH Page 12

47 3.1.1 Lane Change Drivers began Drive 1a in the left lane of RR 27 (the right lane was blocked) and they had to switch the right lane to get to the Highway 406 North ramp. The first lane designation sign was located 145 m west of the Farr Road intersection, and was legible in the simulation 104 m away. If drivers started the change before the Farr Road intersection, they were scored as changing in Zone 1. If they did it after Farr Road but before the tunnel it was scored as Zone 2, and in the Tunnel as Zone 3. As can be seen in Table 5, 2/3 of drivers started their lane changes in Zone 1 east of the Farr Road intersection; one third crossed after the intersection before the tunnel in Zone 3. None made the lane change in the tunnel. In the vast majority of cases, drivers started and completed their lane changes in the same zone. Table 5: Zone in which drivers adopt appropriate lane Zone Frequency Percent Total (Note: Zone = 1 means that drivers take advantage of signage to make the change into the appropriate lane early) Figure 8 shows the distribution of the start and end of lane changes as measured from the tunnel entrance. Note that Farr Road was located 800 m from the tunnel entrance. The closer the distance to 0, the closer the lane change is to the tunnel. UNIVERSITY OF GUELPH Page 13

48 Drive 1a: Percentage of drivers starting and ending lane changes at different distances from the tunnel % Lane change starts Lane change ends Distance from the tunnel (meters) that the change occurs Figure 8: Location of lane changes relative to tunnel entrance Table 6 shows minimum, maximum, and mean distance away from the tunnel that the lane change starts. On average, drivers started their lane change 1052 m from the tunnel and completed the lane change 125 m. Table 6: Location of lane changes relative to tunnel entrance N Minimum Maximum Mean Std. Deviation Lane Change Start Lane Change End Velocity Figure 9 and Table 7 show the velocity profile for drivers in Zones 1, 2, 3 and 4. In Zone 1 east of the Farr Road intersection speeds averaged 80 km/h. It should be noted that all drivers approached the Farr Road intersection signal on a green light. Speed remained constant on the tunnel approach and in the tunnel (Zones 2 and 3). Speed dropped in Zone 4 by about 16 km/h, UNIVERSITY OF GUELPH Page 14

49 likely in response to the abrupt change in scene after leaving the tunnel, and the short distance prior to the curved on-ramp to Hwy 406 N. Drive 1a: Velocity in kph kph Average velocity 10th percentile 25th percentile 75th percentile 90th percentile Zone Figure 9: Velocity profile by zone UNIVERSITY OF GUELPH Page 15

50 Table 7: Velocity profile by zone N Minimum Maximum Mean Std. Deviation Zone Zone Zone Zone Wayfinding Errors Drivers were required to take the bypass ramp in order to reach their destination. By the time drivers reached the bypass ramp they had seen two overhead lane direction signs and one ground mounted sign at the gore. Of the 48 drivers, 46 correctly took the bypass ramp. One driver took the roundabout to get to Highway 406 instead of taking the ramp. One driver did not reach their destination and took the roundabout to Welland instead of taking the bypass to 406, because they forgot their assigned destination. 3.2 Ramp Drive 2a: North on Farr Road, West on RR 27, Ramp to Hwy 406 North An illustration of Ramp Drive 2a is shown below in Figure 10. Figure 10: Ramp Drive 2a including speed measurement zones UNIVERSITY OF GUELPH Page 16

51 Guide signs encountered on Ramp Drive 2a are shown below in Table 8. The first sign encountered on the northbound Farr Road approach to Regional Road 27 was located 250 m from the intersection and was legible at a distance of 82 m; at an approach speed of 80 km/h this sign was legible 14.9 seconds prior to reaching the intersection. The second sign encountered on the northbound Farr Road approach to Regional Road 27 was located 70 m from the intersection and was legible at a distance of 55 m; at an approach speed of 80 km/h this sign was legible for 4.3 seconds prior to reaching the intersection. The first sign west of Farr Road was located 613 m from the tunnel entrance and was legible at a distance of 104 m; at a speed of 70 km/h this sign was legible 36.9 seconds prior to reaching the tunnel entrance. The second sign west of Farr Road was located at the tunnel entrance and was legible at a distance of 101 m; at a speed of 70 km/h this sign was legible 5.2 seconds prior to reaching the tunnel entrance. The first sign west of the canal was located at the gore of the Hwy 406 N bypass and was legible at a distance of 120; at a speed of 66 km/h this sign was legible 6.5 seconds prior to reaching the decision point. Table 8: Guide signs encountered on Ramp Drive 2a Ramp Drive 2a Distance from Decision Point (m) Legibility Distance (m) Speed (km/h) Seconds Legible From Decision Point (sec) UNIVERSITY OF GUELPH Page 17

52 3.2.1 Lane Change Figure 11 shows the distribution of the start and end of lane changes as measured from the tunnel entrance. Note that Farr Road was located 800 m from the tunnel entrance. Note that a negative value indicates that the lane change was completed in the tunnel. Drive 2a: Percentage of drivers starting and ending lane changes at different distances from tunnel % Start lane change End lane change Distance from the tunnel in meters (note that negative values are inside of tunnel) -100 Figure 11: Location of lane changes relative to tunnel entrance The average distance at which lane changes were started was 757 m from the tunnel; the average lane change distance was 89 m (see Table 9). Table 9: Location of lane changes relative to tunnel entrance N Minimum Maximum Mean Std. Deviation Lane Change Start Lane Change End UNIVERSITY OF GUELPH Page 18

53 3.2.2 Velocity Profile Figure 12 and Table 10 show the velocity profile for drivers in Zones 1, 2, 3 and 4. Drive 2a: Velocity in kph kph Average velocity 10th percentile 25th percentile 75th percentile 90th percentile Zone Figure 12: Velocity profile by zone Table 10: Velocity profile by zone N Minimum Maximum Mean Std. Deviation Zone Zone Zone Zone UNIVERSITY OF GUELPH Page 19

54 3.2.3 Bypass Ramp Decision One driver (2.1% of the sample) skipped the bypass ramp and got to Highway 406 using the roundabout. This same driver made it to their destination, though not in the required way. (Note: This driver did exactly the same thing in Drive 1a. They managed to avoid all the bypasses.) Wayfinding Errors One driver (2.1% of the sample) took the roundabout to Welland instead of taking the bypass to the Welland (They did not arrive at the correct destination). 3.3 Roundabout Drive 1b: West on RR 27 to Welland An illustration of Roundabout Drive 1b is shown below in Figure 13. Figure 13: Roundabout Drive 1b, including speed measurement zones The guide signs encountered on Roundabout Drive 1b are shown below in Table 11. UNIVERSITY OF GUELPH Page 20

55 Table 11: Guide signs encountered on Roundabout Drive 1b Roundabout Drive 1b Distance from Decision Point (m) Legibility Distance (m) Speed (km/h) Seconds Legible From Decision Point (sec) Lane change Drivers started east of Farr Road in the right lane because a vehicle was parked in the left lane and they were required to change to get into the left lane to take the roundabout to go to Welland. The distribution of lane change locations is shown below in Table 12. Nearly all drivers (93.8%) made their lane change in Zone 2 (i.e., between Farr Road and the tunnel entrance). UNIVERSITY OF GUELPH Page 21

56 Table 12: Zone in which drivers adopt appropriate lane Zone Frequency Percent Total Figure 14 shows the distribution of the start and end lane changes as measured from the tunnel entrance. Drive 1b: Percentage of drivers starting and end lane changes at different distances from the tunnel % Lane change starts Lane change ends Distance from the tunnel in meters Figure 14: Location of lane changes relative to tunnel entrance On average, drivers started their lane change 800 m from the tunnel entrance and completed their lane change after travelling 133 m (see Table 13). UNIVERSITY OF GUELPH Page 22

57 Table 13: Location of lane changes relative to tunnel entrance N Minimum Maximum Mean Std. Deviation Lane Change Start Lane Change End Velocity profile There was some slowing as drivers approached the bypass ramp (Zone 4). Figure 15 and Table 14 show the velocity profile of subjects as they approached the roundabout. Figure 15: Velocity profile on approach to roundabout UNIVERSITY OF GUELPH Page 23

58 Table 14: Velocity profile on approach to roundabout N Minimum Maximum Mean Std. Deviation zon1vel zon2vel zon3vel zon4vel sp200mrdb sp175mrdb sp150mrdb sp125mrdb sp100mrdb sp75mrdb sp50mrdb sp25mrdb avgspdrdb Collisions in Roundabout Out of 47 drivers who entered the roundabout, one had an at-fault collision within 10 metres of the yield line (2.1% of sample) Lane Use in Roundabout An illustration depicting checkpoints used to evaluate driver lane position in the roundabout is shown in Figure 6. Driver lane position was observed at three locations; roundabout entrance, Checkpoint 4, roundabout exit. The distribution of driver lane selection by location is shown below in Figure 16. UNIVERSITY OF GUELPH Page 24

59 Figure 16: Driver lane choice in roundabout Table 15 below shows what drivers did in the roundabout. In Drive 1b, both entry lanes lead to the destination and thus there was no need for a lane change. Nonetheless, of the 46 drivers who correctly entered the roundabout, 14 made one lane change (30%) in the roundabout, and three of these drivers (3/14 = 21%) missed their exit. Table 15: Driver behaviour in the roundabout Total (n=46) Made no lane unnecessary lane changes 32 Made 1 necessary lane change n/a Made 1 unnecessary lane change 11 Made 1 unnecessary lane change and missed the exit Wayfinding Errors In Drive 1b, drivers come up from the tunnel and are required to drive past the bypass ramp and take the roundabout to Welland. Of the 48 drivers, 46 took the correct path to their destination. One driver (2.1% of sample) took the bypass ramp instead of the roundabout. One driver (2.1% of sample) turned left at the roundabout entrance and circled the roundabout backwards. UNIVERSITY OF GUELPH Page 25

60 3.4 Roundabout Drive 2b: North on Farr Road, West on RR 27 to Welland An illustration depicting Roundabout Drive 2b is shown below in Figure 17. Figure 17: Roundabout Drive 2b including speed measurement zones The guide signs encountered on Roundabout Drive 2b are shown below in Table 16. UNIVERSITY OF GUELPH Page 26

61 Table 16: Guide signs encountered on Roundabout Drive 2b Roundabout Drive 2b Distance from Decision Point (m) Legibility Distance (m) Speed (km/h) Seconds Legible From Decision Point (sec) UNIVERSITY OF GUELPH Page 27

62 3.4.1 Lane Change In this drive, drivers start out northbound on Farr Road in the right lane because the left lane is blocked. Ultimately they have to switch to get into the left lane to get to the appropriate lane for the roundabout. All drivers turned left as required to get on regional road 27. However, a number of drivers actually started their lane change in the intersection before the start of Zone 2 (Zone 2 starts on 27 just west of the intersection at 802 metres from the tunnel.) Because many drivers did the lane change in the intersection (which is in neither Zone 1 nor Zone 2), there are a number that have lane change distances that are very small (and sometimes 0) and similarly small lane change times. Nearly all drivers (97.8%) made their lane change in Zone 2 (see Table 17). One participant made their lane change in Zone 3 (i.e., in the tunnel). Table 17: Zone in which drivers adopt appropriate lane Zone Frequency Valid Percent Total (Note: Zone 3 is in the tunnel.) Figure 18 shows the distribution of the start and end lane changes as measured from the tunnel entrance. Figure 18: Location of lane changes relative to tunnel entrance UNIVERSITY OF GUELPH Page 28

63 On average, drivers started their lane change 773 m from the tunnel entrance and completed their lane change after travelling 32 m (see Table 18). Table 18: Location of lane changes relative to tunnel entrance Lane Change Start Lane Change End N Minimum Maximum Mean Std. Deviation Velocity Profile The velocity profile of drivers on approach to the roundabout is shown below in Figure 19 and Table 19. Figure 19: Velocity profile on approach to roundabout UNIVERSITY OF GUELPH Page 29

64 Table 19: Velocity profile on approach to roundabout N Minimum Maximum Mean Std. Deviation zon1vel zon2vel zon3vel zon4vel sp200mrdb sp175mrdb sp150mrdb sp125mrdb sp100mrdb sp75mrdb sp50mrdb sp25mrdb avgspdrdb Collisions in Roundabout Of the 47 drivers who entered the roundabout, one driver had an at-fault collision at the roundabout entrance (2.1% of sample) Lane Use in Roundabout In Drive 2b, drivers were required to enter the roundabout, pass one exit (and checkpoint 4) and proceed to the second exit in order to reach Welland (see Figure 6). Vehicle lane positions were recorded at the roundabout entrance, at checkpoint 4 and at the roundabout exit (see Figure 20). UNIVERSITY OF GUELPH Page 30

65 Figure 20: Driver lane choice in roundabout Table 20 shows what drivers did in the roundabout. In Roundabout Drive 2b, both entry lanes lead to the destination and thus there was no need for a lane change. Nonetheless, of the 46 drivers who entered the roundabout, 14 made between one and three lane changes (30%) in the roundabout, and three of these drivers (21%) missed their exit. Table 20: Driver behaviour in the roundabout Total (n=46) Made no lane unnecessary lane changes 32 Made 1 necessary lane change n/a Made 1 unnecessary lane change 10 Made 1 unnecessary lane change and missed their exit 1 Made 2 unnecessary lane changes 1 Made 3 unnecessary lane changes and missed their exit Wayfinding Errors Of the 48 drivers, five (10% of sample) experienced difficulty reaching their destination. The following wayfinding errors were observed: One driver went the wrong way at the Farr Road intersection (2.1% of sample) One driver turned into oncoming traffic at the Farr Road intersection and continued on through the tunnel. On the other side, when the median cleared, the participant jumped UNIVERSITY OF GUELPH Page 31

66 the centre island, returned to the roundabout and ended up in an expected location (2.1% of the sample) Three drivers took the wrong exit in the roundabout (6.3% of the total sample of 48) 3.5 Roundabout Drive 3: East on Main St to Hwy 406 North An illustration depicting Roundabout Drive 3 is shown below in Figure 21. Figure 21: Roundabout Drive 3 including speed measurement zones The guide signs encountered on Roundabout Drive 3 are shown below in Table 21. UNIVERSITY OF GUELPH Page 32

67 Table 21: Guide signs encountered on Roundabout Drive 3 Sign 1 Roundabout Drive 3 Distance from Decision Point (m) Legibility Distance (m) Speed (km/h) Seconds Legible From Decision Point (sec) Sign Sign Sign Velocity Profile The velocity profile of drivers on approach to the roundabout is shown below in Figure 22 and Table 22. UNIVERSITY OF GUELPH Page 33

68 Drive 3: Velocity in kph kph z Zone and distance to roundabout in meters (0 means in the roundabout) Average velocity 10th percentile 25th percentile 75th percentile 90th pecentile Figure 22: Velocity profile on approach to roundabout Table 22: Velocity profile on approach to roundabout N Minimum Maximum Mean Std. Deviation sp200mrdb sp175mrdb sp150mrdb sp125mrdb sp100mrdb sp75mrdb sp50mrdb sp25mrdb avgspdrdb Collisions in Roundabout Of the 48 drivers who entered the roundabout three drivers had an at-fault collision in the area of the roundabout (6.2% of sample); two drivers had collisions at the roundabout entrance and one driver had a collision 11 metres inside the roundabout. UNIVERSITY OF GUELPH Page 34

69 3.5.3 Lane Choice in Roundabout In Drive 3, drivers travelled east on the Regional Road 27 toward the roundabout and then were required to take the roundabout to Highway 406. This would involve passing checkpoints 2 and 3 in the roundabout (see Figure 6). Driver lane position was observed at four locations: roundabout entrance, Checkpoint 2, Checkpoint 3, and roundabout exit. The distribution of driver lane selection by location is shown below in Figure 23. Drivers were required to be in lanes 1 or 2 in order to reach the Hwy 406 exit. At the roundabout entrance 22% of drivers were in the wrong lane, at Checkpoint 2 about 10% of drivers were in the wrong lane, and at Checkpoint 3 all drivers had moved out, or were in the process of moving out of the incorrect lane. Figure 23: Driver lane choice in roundabout Table 23 shows what drivers did in the roundabout. Of 48 drivers, 37 entered in the correct lane. Of the 37 drivers who entered the roundabout in the correct lane 19 drivers did not make any lane changes (51%) and 18 drivers made unnecessary lane changes or took the wrong exit (49%); 10 out of 18 drivers made one unnecessary lane change (27%); two out of 18 drivers made two unnecessary lane changes (5%); three out of 18 drivers made three unnecessary lane changes (8%); two out of 18 drivers were in between lanes at one of the checkpoints (5%); one driver missed their exit (3%). Of the 11 drivers that were in the wrong lane at the entrance, eight drivers correctly changed lanes to get into the appropriate lane, one driver made an extra lane change, and two drivers missed their exit. UNIVERSITY OF GUELPH Page 35

70 Table 23: What drivers did in the roundabout Researcher Observations Entered roundabout in correct lane (n=37) Entered roundabout in wrong lane (n=11) Total (n=48) Made no lane unnecessary lane changes Made 1 necessary lane change N/A 8 8 Made 1 unnecessary lane change Made 2 unnecessary lane changes Made 3 unnecessary lane changes Were between lanes at checkpoint Missed their exit Wayfinding Errors Drivers were required to enter the roundabout and take the second exit to reach Hwy 406 N. Nearly all drivers (95.7%) took the correct exit in the roundabout. Two drivers took the wrong exit in the roundabout (see Table 23 above). For drivers who took the correct exit in the roundabout, nearly all (97.8%) took the correct exit on their first lap in the roundabout. One driver had to go around the roundabout and took the correct exit on the second encounter (see Table 23 above). 3.6 Roundabout Drive 4: North on Wellington, East on Main St to Hwy 406 North An illustration depicting Roundabout Drive 4 is shown below in Figure 24. Figure 24: Roundabout Drive 4, including speed measurement zones UNIVERSITY OF GUELPH Page 36

71 The guide signs encountered on Roundabout Drive 4 are shown below in Table 24. Table 24: Guide signs encountered on Roundabout Drive 4 Roundabout Drive 4 Distance from Decision Point (m) Legibility Distance (m) Speed (km/h) Seconds Legible From Decision Point (sec) Velocity Profile The velocity profile of drivers on approach to the roundabout is shown below in Figure 25 and Table 25. UNIVERSITY OF GUELPH Page 37

72 Drive 4: Velocity in kph kph Average velocity 10th percentile 25th percentile 75th percentile 90th percentile Distance to roundabout in meters Figure 25: Velocity profile on approach to roundabout Table 25: Velocity profile on approach to roundabout N Minimum Maximum Mean Std. Deviation sp175mrdb sp150mrdb sp125mrdb sp100mrdb sp75mrdb sp50mrdb sp25mrdb avgspdrdb Collisions in Roundabout Of the 48 drivers who entered the roundabout, no drivers had a collision Lane Choice in Roundabout In drive 4 drivers travelled north on Wellington and turned right onto East Main St. They were required to go through the roundabout and take the exit for Highway 406. This would involve passing checkpoints 2 and 3 in the roundabout (see Figure 6). UNIVERSITY OF GUELPH Page 38

73 Vehicle lane positions were recorded at the roundabout entrance, at Checkpoint 2, Checkpoint 3, and at the roundabout exit (see Figure 26). Drivers were required to be in lanes 1 or 2 in order to reach the Hwy 406 exit. At the roundabout entrance approximately 30% of drivers were in the wrong lane and at Checkpoint 2 only 10% of drivers were in the wrong lane. By Checkpoint 3 all drivers were either in the process of moving out of the outside lane, or had already moved to lane 1 or 2. Figure 26: Driver lane choice in roundabout Table 26 shows what drivers did in the roundabout. Of 48 drivers, 33 entered the roundabout in a correct lane (69%) and 15 entered in the incorrect lane (31%). Overall, 14 out of 48 drivers made unnecessary lane changes while in the roundabout (29%) UNIVERSITY OF GUELPH Page 39

74 Table 26: What did the drivers do? Researcher Observations Entered roundabout in correct lane (n=33) Entered roundabout in wrong lane (n=15) Made no unnecessary lane changes Made 1 necessary lane change N/A 1 1 Made 1 unnecessary lane change Made 2 unnecessary lane changes Made 3 unnecessary lane changes Were between lanes at checkpoint Missed their exit Wayfinding Errors Total (n=48) Most drivers (83.3%) took the correct exit to access Hwy 406. Eight drivers (16.7%) took the wrong exit (see Table 26 above). All drivers (100%) took the correct exit on their first lap in the roundabout. 3.7 Roundabout Drive 5: South on Hwy 406, take roundabout East to RR 27 An illustration depicting Roundabout Drive 5 is shown below in Figure 27. Figure 27: Roundabout Drive 5, including speed measurement zones UNIVERSITY OF GUELPH Page 40

75 The guide signs encountered on Roundabout Drive 5 are shown below in Table 27. Table 27: Guide signs encountered on Roundabout Drive 5 Roundabout Drive 5 Distance from Decision Point (m) Legibility Distance (m) Speed (km/h) Seconds Legible From Decision Point (sec) Velocity Profile The velocity profile of drivers on approach to the roundabout is shown below in Figure 28 and Table 28. UNIVERSITY OF GUELPH Page 41

76 Drive 5: Velocity in kph kph Average velocity 10th percentile 25th percentile 75th percentile 90th percentile 10 z Zone and distance to roundabout in meters (0 means in roundabout) Figure 28: Velocity profile on approach to roundabout Table 28: Velocity profile on approach to roundabout N Minimum Maximum Mean Std. Deviation 2 km away km away zon6vel sp200mrdb sp175mrdb sp150mrdb sp125mrdb sp100mrdb sp75mrdb sp50mrdb sp25mrdb avgspdrdb UNIVERSITY OF GUELPH Page 42

77 3.7.2 Collisions in Roundabout Of the 48 drivers who entered the roundabout, no drivers had a collision Lane Choice in Roundabout Drivers traveled south on Hwy 406 and were required to travel east on Regional Road 27 toward Port Colborne. This required entering the roundabout and passing Checkpoints 1 and 2 (see Figure 6). Vehicle lane positions were recorded at the roundabout entrance, at Checkpoint 1, Checkpoint 2, and at the roundabout exit (see Figure 29). Drivers were required to be in lanes 2 or 3 in order to reach the Regional Road 27 exit. Figure 29: Driver lane choice in roundabout Table 29 shows what drivers did in the roundabout. In Roundabout Drive 6 both entry lanes could be used to reach the destination thus no lane changes were required. Of the 44 drivers that entered the roundabout 30 changed lanes unnecessarily or drove between lanes (68%). Overall, only one driver took the wrong exit. UNIVERSITY OF GUELPH Page 43

78 Table 29: Driver behaviour in the roundabout Total (n=44) Made no lane unnecessary lane changes 14 Made 1 necessary lane change n/a Made 1 unnecessary lane change 13 Made 2 unnecessary lane changes 12 Made 3 or more unnecessary lane changes 3 Driving between lanes Wayfinding Errors In Roundabout Drive 6 drivers were required to use the roundabout, and not the slip lane, to reach Welland. Of the 48 drivers, five (10% of sample) experienced difficulty reaching their destination. The following wayfinding errors were observed: Three took the ramp or slip lane and missed the roundabout (6.3% of the sample) One driver (2.1% of the sample) turned left and took the roundabout backwards and never left (and thus did not achieve the destination). One took the wrong exit in the roundabout (2.1% of the sample) For drivers who correctly entered the roundabout, most (83.7%) took the correct exit on their first lap in the roundabout. Seven drivers (16.3%) had to go around the roundabout and took the correct exit on the second lap (see Table 29 above). 3.8 Practice Effects Across Similar Drives There are a very large numbers of ways that this data could be analyzed, and many different combinations of orders of presentation for the drives where practice effects might be evident. (The drivers did, after all do 5 roundabouts, and though these roundabouts started and ended at different points, early experience with any roundabout might have an effect.) There are also a very large number of dependent measures. Consequently, this is not a complete or exhaustive listing of the practice effects that could be examined, but rather a sampling. Nonetheless, it is useful to consider the effect of doing two drives that are identical at least up to a point, to see if the drivers behave differently on a given segment of road based on experience based on exposure to an identical stretch of that same highway Location of the Lane Change in Drives 1a and 1b Drives 1a and 1b start at the same place (east of Farr Road) and both go down 27 though 1a goes to a ramp and 1b goes to a roundabout. In both cases drivers have to make a lane change in response to information on a sign, but they will be changing into different lanes. We are looking at the effects of previous experience on the similar drive has on the place the lane change starts and stops. Thus, we are comparing the distance at which the lane change occurred in Drive 1a depending on whether Drive 1a was done first (before Drive 1b) or second. There was no significant effect of having experience with Drive 1b on lane change distance for Drive 1a. However, there was a significant effect of experience with Drive 1a on both the position that the lane change began and ended in Drive 1b (p =.02 and p =.04) respectively, UNIVERSITY OF GUELPH Page 44

79 with the lane change occurring earlier as a result of experience with Drive 1a. Drivers began and ended their lane change earlier as a result of experience (see Figure 30). Figure 30: Location of lane change manoeuvres Drive 1a and 1b Drive 2a and 2b are also similar, insofar as they both started south of Regional Road 27 on Farr Road both involved a left turn and the need to get into appropriate lane thereafter. However, in this case, there were no notable differences in the position of the lane change as a function of practice with a similar drive (see Figure 31). UNIVERSITY OF GUELPH Page 45

80 Figure 31: Location of lane change manoeuvres Drive 2a and 2b Speed Reductions on Roundabout Approach and Number of Stops Roundabout Drives 1b and 2b both involve going to the roundabout, though the two drives differ in terms of where they start (east of Farr as opposed to South on Farr). When it comes to Drive 1b, the speed profile to the roundabout was about the same whether or not there was previous experience with Drive 2b. There was a reduction in the number of stops at the roundabout as a function of whether Drive 1b was first or second (M =.42 to.25). There were no significant differences in speed at any point (see Figure 32). For Roundabout Drive 2b, there were marginal decreases in speed as a result of experience with 1b at 100 metres and 25 metres (p =.10 and p =.09). There was also a marginally significant increase in the number of stops at the entrance to the roundabout after experience with Ramp Drive 2a (M =.18 to.42, p =.09). UNIVERSITY OF GUELPH Page 46

81 Figure 32: Practice effects on velocity Drive 1b and 2b Roundabout Drives 3 and 4 both involved travelling east along 27 to arrive at the roundabout, though the drives differed in their starting point (East Main vs. Wellington) but for both the goal was Highway 406. The driving speeds did not differ significantly as a function of whether drivers had previous experience with the similar section of highway before. See graph on the following page for the speed profile on the lead up to the roundabout. Note that for Roundabout Drive 4, drivers are turning right off Wellington and there is not as much space between the starting point and the roundabout (there is no measure for 200 metres as a result) (see Figure 33). UNIVERSITY OF GUELPH Page 47

82 Figure 33: Practice effects on velocity Drive 3 and 4 4 DRIVER SIGN QUESTIONNAIRE 4.1 Immediate Recognition Test Following the simulator drives participants were shown one at a time, on paper, 33 signs, including 21 signs they had seen in the simulator as well as 12 distractor signs, but which they had not seen but were similar. The 33 signs were spread out over the surface of a large table, and the order that the signs appeared was randomized for each participant. Participants were told that the table held some signs that had been presented on their simulated drive (the targets), and some that had not (the distractors). They were asked to pick out the target signs. Participants were given no indication of the number of targets or distractors. They were asked to pick out the signs they had seen on the simulated drives and later asked to rate their level of confidence, on a nine-point scale, that they in fact saw the sign. A test of which signs were remembered best indicated that the ground-mounted, regulatory lane designation signs were not well remembered. Subjects also rated their confidence that they had seen these signs in the simulation as low. Both findings suggest that drivers did not allocate much attention to these signs. They are visible at the same time as the overhead lane designation sign and therefore compete for the subject s attention. Because they are ground mounted and to the sides of the roadway, it is not as clear which lane is referred to as is the case for the overhead lane designation signs. 4.2 Sign Questionnaire Following the Immediate Recognition Test participants were shown each sign on a piece of paper and were asked the following questions: UNIVERSITY OF GUELPH Page 48

83 1. Ease of Reading: This is a question about how easy this sign would be to read. If this sign appeared on a highway, do you think that people would generally find it easy to read from a distance? (on a scale of 1 (extremely hard) to 9 (extremely easy)) 2. Ease of Understanding: This is a question about how easy this sign would be to understand. If this sign appeared on a highway, do you think that other people would generally find it easy to understand? (on a scale of 1 (extremely difficult) to 9 (extremely easy)) 3. How Well Sign Prepares Driver: How good a job do you think this sign does in preparing drivers for what is coming up ahead on the road? (on a scale of 1 (extremely bad) to 9 (extremely good)) 4. What is the meaning of the sign? Conventional guide signs performed best and roundabout regulatory signs, worst (see Tables 30, 31, 32, and 33 below). In the analysis that follows, an asterisk indicates that the score is significantly greater than the midpoint rating (5 on a 9 point class; one-sample t-tests, p <.05). An analysis of variance revealed that on average, roundabout-related signs were significantly harder to read and understand, and were judged as significantly worse preparation than the other signs (p <.001). Post-hoc tests revealed the same pattern emerged when comparing each of the individual roundabout signs against the average for the conventional signs in terms of ease of reading and understanding. Table 30: Comparison between sign types Ease of Sign reading Standard Guide Signs 8.1* (0.2) Roundabout Guide Signs 7.3* (0.1) Roundabout Regulatory Signs 7.1* (0.5) Note: (1 = extremely bad, 9 = extremely good) Ease of understanding 8* (0.2) 6.6* (0.2) 6.1* (0.7) How well sign prepares driver 7.8* (0.4) 6.3* (0.4) 5.8* (0.8) UNIVERSITY OF GUELPH Page 49

84 Table 31: Roundabout guidance category all signs Code Sign Ease of reading 1 7.4* (1.4) Ease of understanding 6.7 * (1.8) How well sign prepares driver 6.4* (1.7) 3 7.3* (1.7) 6.3* (2.0) 5.9* (2.) 5 7.4* (1.6) 6.4* (1.9) 5.9* (2.1) 6 7.3* (1.6) 6.9* (1.9) 7.1* (1.5) * (1.5) 6.6* (1.7) 6.2* (1.7) TOTAL 7.3* (0.1) 6.6* (0.2) 6.3* (0.4) UNIVERSITY OF GUELPH Page 50

85 Table 32: Roundabout regulatory category all signs Code Sign Ease of reading * (1.3) Ease of understanding 7.5* (1.6) How well sign prepares driver 7.4* (1.7) * (1.8) 5.4 (2.0) 4.9 (2.2) * (1.9) 5.9* (1.8) 5.8* (2.0) (1.7) 6.2 * (2.2) 6.1* (2.2) * (1.7) 6.0* (2.1) 5.8* (1.9) * (2.1) 5.4 (2.2) 4.9 (2.2) * (1.7) 6.4 * (1.9) 5.9* (2.0) TOTAL 7.1* (0.5) 6.1* (0.7) 5.8* (0.8) UNIVERSITY OF GUELPH Page 51

86 Table 33: Standard Guidance Category All Signs Code Sign Ease of reading 2 8.0* (1.4) Ease of understanding 8.0 * (1.6) How well sign prepares driver 8.0* (1.7) * (1.2) 7.8* (1.6) 7.3* (2.1) * (1.2) 8.1* (1.6) 8.0* (1.6) TOTAL 8.1* (0.2) 8* (0.2) 7.8* (0.4) When asked about the meaning of the sign, for the roundabout signs with poor comprehension scores, a common error was to assume that they simply meant that there would be a curve in the road ahead. In their comments, a few drivers commented that the presence of the dot in the centre clarified the meaning to some extent, but others indicated that they did not understand the significance of the dot. Dutch (hook-style) roundabout signs were posted in the lab from a previous iteration of the study, and some of these were used as distractor items in the immediate recognition test described earlier. It should be noted that the hook-style signs give a stronger impression of the tight radius to be encountered than the signs used in the study. Some of the drivers commented that the Dutch-style signs were easier to understand than the ones used in the study. Consideration should be given to testing these in a future study. Age had a significant effect on sign comprehension scores. Comprehension scores were significantly higher for the youngest age group (18-24), followed by the medium and older age groups (p <.05 for each). 4.3 V Scope Reaction Time Testing A PC-based program was created by taking static screenshots of road signs embedded in the road scene from the driving simulator. Participants were told that their task would to be to make a series of decisions about which lane to take to arrive at specific destinations. Participants were seated in front of a computer terminal, and on the keyboard there were three response keys: a key labelled with a left arrow (to indicate that the driver should be in the left lane to get to the specified destination); a key labelled with a right arrow (to indicate the driver should be in the right lane to get to the destination); and an arrow pointing straight up (indicating that the driver should go into the middle lane to get to the destination). The middle arrow option was only a possibility on a small subset of the trials, and the question display cued drivers to when this would be a possibility. After practice trials, participants did a series of experimental trials in which they were presented with a destination (e.g. Which lane to St. Catharines? ), then shown UNIVERSITY OF GUELPH Page 52

87 a scene from the drive, and asked to indicate their decision about which lane to take by pushing the left, right, or middle lane response key. The question was presented for one second and then driving scene was shown. The scene remained on screen until the participant made a response by pushing a response key. If the participant did not make a response within 4 seconds, the program timed out and went to the next trial. In an earlier study of bilingual variable message signs, 4 seconds was used as the display time period based on a calibration study carried out on the road and in the laboratory showing that subject sign comprehension performance with this time period in the laboratory matched that on the road (Smiley et al. 2005). The 4 second timeout period is more than twice as long as the average reaction time in this task. Sign images were presented in a random order so that every participant received the trials in a different order. There were a total of 23 trials. Forty-five drivers completed this portion of the study. Response times measured the time between the presentation of the driving scene and the participants key press response. Statistical analyses were carried out, first by comparing the proportion of participants who could not make a decision with 4 seconds in each condition with the average proportion across conditions in (Chi square goodness of fit). The second analysis involved comparing the average decision for each situation for the overall average latency across conditions (M = 1994 ms). Overall, the best performing signs with respect to average response time and the percent of timeout trials were conventional overhead guide signs on the westbound approach. Overhead roundabout guide signs performed better than ground-mounted map guide signs. With the one exception of drivers taking the slip lane to Welland to the right, response times to right destinations were faster than those for left or through destinations. The reduced time for right destinations (the first choice point) may reflect the greater wayfinding effort involved in using a roundabout to reach through and left destinations especially for unfamiliar drivers. The average response time to determine lane choice was significantly longer when there were two correct alternatives as compared to one (p <.001, n = 45). In addition, there was a significantly higher proportion of timeout trials when there were two correct alternatives as compared to when there was only one (26.7% vs. 11.1%, p <.001) (see Figure 34). UNIVERSITY OF GUELPH Page 53

88 Figure 34: Proportion of timeout trials as a function of number of correct alternatives A detailed analysis is included in the following sections Westbound Approach Participants were shown the sign below, located on Farr Road, and asked which lane they should take to get to St. Catharines (see Table 34). The mean response time was 2.20 seconds and 89% of participants correctly identified the left lane. Four of 44 participants did not select a lane within the 4.0 second period. The mean response time for this sign was greater than the overall mean response time (2.22 sec vs sec, p =.022). The proportion of participants who timed out (i.e., did not respond with 4.0 seconds) did not differ from the overall mean. UNIVERSITY OF GUELPH Page 54

89 Table 34: Facing northbound drivers, guide sign on Farr Road Destination: St. Catharines N % Sample Mean Response Time (SD) Correct Lane msec (738) Incorrect Lane 0 0 Timeout Total Participants were shown the sign below on two separate occasions (see Table 35). On one viewing they were asked which lane they should use to reach Welland, and on a separate viewing they were asked which lane they should use to reach St. Catharines. The mean response time for the Welland destination was significantly faster than the overall mean (1.41 sec vs sec, p <.001) and the proportion of participants who timed out before responding was significantly less than average (2.2% vs. 15.4%, p <.05). The mean response time for the St. Catharines destination (1.74 sec) was not different than the overall mean and the proportion of participants who timed out before responding (6.7%) was not significantly different than the average proportion across all signs. UNIVERSITY OF GUELPH Page 55

90 Table 35: Facing westbound drivers, first guide sign west of Farr Rd Destination: Welland Destination: St. Catharines N % Sample Mean Response Time (SD) N % Sample Mean Response Time (SD) Correct Lane (658) (828) Incorrect Lane Anticipation, RT < 125 Timeout Total Participants were shown the sign below on two separate occasions (see Table 36). On one viewing they were asked which lane they should use to reach Welland, and on a separate viewing they were asked which lane they should use to reach St. Catharines. The mean response time for the Welland destination was significantly faster than the overall mean (1.43 sec vs sec, p <.001) and the proportion of participants who timed out before responding (4.4%) was not different than the overall average. The mean response time for the St. Catharines destination was significantly faster than the overall mean (1.35 sec vs sec, p <.001) and the proportion of participants who timed out before responding was significantly less than the average proportion across all signs (0% vs. 15.4%, p <.05). UNIVERSITY OF GUELPH Page 56

91 Table 36: Facing westbound drivers, second guide sign west of Farr Rd Destination: Welland Destination: St. Catharines N % Sample Mean Response Time (SD) N % Sample Mean Response Time (SD) Correct Lane 39* (691) (652) Incorrect Lane (894) Anticipation Timeout Total Participants were shown the sign below on two separate occasions (see Table 37). On one viewing they were asked which lane they should use to reach East Main, and on a separate viewing they were asked which lane they should use to reach St. Catharines. The mean response time for the East Main destination was significantly slower than the overall mean (2.41 sec vs sec, p <.01) and the proportion of participants who timed out before responding was greater than the overall average (44.4% vs. 15.4%, p <.05). The mean response time for the St. Catharines destination was significantly slower than the overall mean (2.56 sec vs sec, p <.001) and the proportion of participants who timed out before responding (8.9%) was not different than the average proportion across all signs. UNIVERSITY OF GUELPH Page 57

92 Table 37: Facing westbound drivers, map sign on approach to roundabout Destination: East Main Destination: St. Catharines N % Sample Mean Response Time (SD) N % Sample Mean Response Time (SD) Correct Lane (629) (781) Incorrect Lane Timeout Total Participants were shown the sign below on two separate occasions (see Table 38). On one viewing they were asked which lane they should use to reach Welland, and on a separate viewing they were asked which lane they should use to reach St. Catharines. The mean response time for the Welland destination was significantly longer than the overall mean (2.29 sec vs sec, p <.03) and the proportion of participants who timed out before responding was greater than the overall average (35.6% vs. 15.4%, p <.05). The mean response time for the St. Catharines destination was (1.75 msec) not different than the overall mean and the proportion of participants who timed out before responding (8.9%) was not different than the average proportion across all signs. UNIVERSITY OF GUELPH Page 58

93 Table 38: Facing westbound drivers, lane designation sign on approach to roundabout Destination: Welland Destination: St. Catharines N % Sample Mean Response Time (SD) N % Sample Mean Response Time (SD) Correct Lane (658) (798) Incorrect Lane Timeout Total Southbound Approach Participants were shown the sign below on two separate occasions (see Table 39). On one viewing they were asked which lane they should use to reach Highway 140 and on a separate viewing they were asked which lane they should use to reach East Main Street. The mean response time for the Highway 140 destination was significantly longer than the overall mean (2.27 sec vs sec, p <.03) and the proportion of participants who timed out before responding was greater than the overall average (28.9% vs. 15.4%, p <.05). The mean response time for the East Main Street destination was (2.19 sec) not different than the overall mean and the proportion of participants who timed out before responding (8.9%) was not different than the average proportion across all signs. UNIVERSITY OF GUELPH Page 59

94 Table 39: Facing southbound drivers, first map sign on approach to roundabout Destination: Highway 140 Destination: East Main Street N % Sample Mean Response Time (SD) N % Sample Mean Response Time (SD) Correct Lane (807) (792) Incorrect Lane (343) 0 0 Timeout Total Participants were shown the sign below on four separate occasions (see Table 40). On two viewings they were asked which lane they should use to reach Port Colborne and on two separate viewings they were asked which lane they should use to reach Welland. The mean response time for the Port Colborne destination (2.08 sec) was not different than the overall mean and the proportion of participants who timed out before responding (21.3%) was not different than the overall average. The mean response time for the Welland destination (2.16 sec) was not different than the overall mean and the proportion of participants who timed out before responding (13.3%) was not different than the average proportion across all signs. UNIVERSITY OF GUELPH Page 60

95 Table 40: Facing southbound drivers, second map sign on approach to roundabout Destination: Port Colborne N1 N2 N- Total % Correct Lane Incorrect Lane Mean Response Time (SD) 2084 (842) Destination: Welland N1 N2 N- Total % Mean Response Time (SD) 2160 (827) Timeout Total Participants were shown the sign below on two separate occasions (see Table 41). On one viewing they were asked which lane they should use to reach Port Colborne and on one viewing they were asked which lane they should use to reach Welland. The mean response time for the Port Colborne destination (1.95 sec) was not different than the overall mean and the proportion of participants who timed out before responding (22.2%) was not different than the overall average. The mean response time for the Welland destination was significantly faster than the overall mean (1.73 sec vs sec, p <.03) and the proportion of participants who timed out before responding was significantly less than the average proportion across all signs (2.2% vs. 15.4%, p <.05). UNIVERSITY OF GUELPH Page 61

96 Table 41: Facing southbound drivers, lane designation sign on approach to roundabout Destination: Port Colborne Destination: Welland N % Sample Mean Response Time (SD) N % Sample Mean Response Time (SD) Correct Lane (721) (701) Incorrect Lane (2190) Timeout Total Eastbound Approach Participants were shown the sign below on four separate occasions (see Table 42). On two viewings they were asked which lane they should use to reach St. Catharines and on two separate viewings they were asked which lane they should use to reach Port Colborne. The mean response time for the Port Colborne destination was significantly longer than the overall mean (2.56 vs sec, p <.001) and the proportion of participants who timed out before responding (23.3%) was not different than the overall average. The mean response time for the Welland destination was significantly longer than the overall mean (2.41 vs sec, p <.01) and the proportion of participants who timed out before responding (27.8%) was not different than the average proportion across all signs. UNIVERSITY OF GUELPH Page 62

97 Table 42: Facing eastbound drivers on approach to roundabout Correct Lane Incorrect Lane Destination: St. Catharines N1 N2 N- Tota l % Mean Response Time (SD) (731) (931) Destination: Port Colborne N1 N2 N- Tota l % Mean Response Time (SD) (840) Timeout Total Driver Feedback General Comments on Number and Location of Signs Participants were asked the following three multiple choice questions. The percentage of participants that endorsed each response is listed below the question. 1. Overall, considering all of the drives, check the one response that best expresses how you feel about the NUMBER of signs. There were... Exactly the right number of signs 23% Sometimes too many signs 29% Sometimes too few signs 19% Sometimes too many and sometimes too few signs 29% Total n = 48 UNIVERSITY OF GUELPH Page 63

98 2. If you said there were sometimes too many signs, WHERE were there too many signs? (CHECK ANY THAT APPLY.) Also, if you answered that there were sometimes too many and sometimes too few signs, please fill this in as is appropriate. On the highway a kilometre before the roundabout 19% On the highway immediately before the roundabout 56% In the roundabout 16% On the section of the road, after the roundabout 0 On the highway before the tunnel 6% On the road after the tunnel 0 Total n = If you said there were sometimes too few signs, WHERE were there too few signs? (CHECK ANY THAT APPLY.) Also, if you answered that there were sometimes too many and sometimes too few signs, please fill this in as is appropriate. On the highway a kilometre before the roundabout 50% On the highway immediately before the roundabout 14% In the roundabout 14% On the section of the road, after the roundabout 0 On the highway before the tunnel 21% On the road after the tunnel 0 Total n = Driver Comments on Signs The following is a list of driver comments on the signs used in the simulation Might have been that the signs didn't come into focus soon enough, but I was scrambling in the roundabout Right number of signs Signs too close to 406 exit (turn off) There should be a separate sign for each exit destination More signs to worn of the roundabout and directions in the roundabout before the tunnel More signs 1k before with consistent information as related to signage within or near roundabout would be better approach Good spacing, easy to follow and understand The signs were not clear enough for me Signs that are located overhead of their respective lanes are easier to follow Too close together Speed limits to slow me down would have been good There should have been more notice coming up to the round about I found the signs before roundabouts too close for me to react quick enough to make the right lane choice and to get off at the right direction More advanced warning of which lane you should be in before entering the roundabout In some cases, the signs were placed too close to the decision point. Entering into the roundabout the road Service signs distracted from the direction desired Hard to go on the roundabout UNIVERSITY OF GUELPH Page 64

99 5 DISCUSSION 5.1 Positive Guidance Findings The simulation was organized into a series of seven drives, intended to help determine whether the positive guidance proposed by MTO results in: 1. Westbound drivers on Regional Road 27 heading to Hwy 406 N moving into the right lane well before reaching the tunnel (Drive 1a) 2. Westbound drivers on Regional Road 27 heading to Welland moving into the left lane well before reaching the tunnel (Drive 1b) 3. Northbound drivers on Farr Road turning left and heading to Hwy 406 N moving promptly into the rightmost lane before reaching the tunnel (Drive 2a) 4. Northbound drivers on Farr Road turning left and heading to Welland moving promptly into the leftmost lane before reaching the tunnel (Drive 2b) 5. Eastbound drivers on East Main Street heading to Hwy 406 N moving into the left lane on the approach to the roundabout and turning right as directed within the roundabout (Drive 3) 6. Eastbound drivers entering East Main Street from Wellington Street heading to Hwy 406 N moving into the left lane on the approach to the roundabout and turning right as directed within the roundabout (Drive 4) 7. Southbound drivers on Hwy 406 heading to Regional Road 27 taking either lane on entry to the roundabout (both lanes will take drivers to Regional Road 27) (Drive 5) Following the simulated drives, subjects completed a sign comprehension questionnaire, carried out a V Scope reaction time test to selected signs and provided general feedback on sign effectiveness Westbound & North to Westbound Approaches (Drives 1a, 1b, 2a, 2b) The first question was whether the positive guidance plan was effective in getting drivers heading to Hwy 406N into the right lane well before entering the tunnel. This was tested in Drives 1a (Regional Road 27 to Highway 406 North) and 2a (Farr Road NB to Hwy 406N). On the northbound to westbound approach to the roundabout (Drives 1a and 1b), drivers encountered two ground mounted signs indicating a left turn for the destinations of Welland and Hwy 406 to St. Catharines. Northbound drivers who turned left to go westbound, as well as westbound drivers on Regional Road 27 encountered an overhead lane designation sign indicating Hwy406 N to St. Catharines in the right lane and Welland in the left lane 145 m to the west of the Farr Road intersection. A second overhead lane designation sign was located on the tunnel entrance, 800 m from Farr Road. Figure 35 shows sign layouts, legibility distance of each sign in the simulator (where visual resolution is less than the eye in real life), measured speed in the simulator and legibility distance from point of first legibility to the decision point, at the measured speed. UNIVERSITY OF GUELPH Page 65

100 Figure 35: Guide signs on westbound approach There were a total of 96 trials for drivers heading to Highway 406 North. Each of 48 drivers approached the tunnel twice, once from Regional Road 27 westbound (Drive 1a), and once from Farr Road northbound. (Drive 2a). A stalled vehicle was used to force drivers initially into the wrong lane (left lane) for their destination, so that they would have to change lanes. On Drive 1a, 2/3 of drivers changed lanes back to the right lane immediately after passing the stalled vehicle in Zone 1 before the Farr Road intersection. The guide sign placed 145 m west of the Farr Road intersection could only be read at 104 m away, about 40 m after crossing the Farr Road intersection. Thus most drivers started their lane changes before they could actually read the guide sign either due to a desire to return to the right lane, or as a result of practice on previous drives. On Drive 2a, drivers encountered a stalled vehicle in the left turn lane. All of the drivers went to the right lane, with most lane changes being initiated immediately after the Farr Road intersection, just before most drivers could likely read the guide sign. Thus although most drivers were in the correct lane when entering the tunnel, it seems to have been because they preferred to be in the right lane. UNIVERSITY OF GUELPH Page 66

101 In all but 1 out of 96 trials drivers were fully in the correct lane at least 450 m before the tunnel. The remaining driver drove in the eastbound lanes after turning onto Regional Road 27, into the tunnel, finally returning to the correct side of the road just before the roundabout. In summary, the positive guidance on the westbound approach up to the ramp bypass following the tunnel appears to have worked well. A second positive guidance question was whether drivers would use the bypass lane instead of the roundabout. Mean speed within the tunnel was 83 km/h for both Drives 1a and 2a. With respect to accessing the bypass lane to Highway 406 North, drivers slowed by about 16 km/h (nearly 20%) as they exited the tunnel. This slowing may have been in response to the appearance of the ramp decision point located approximately 50 m beyond the tunnel, equivalent to just over 2 seconds at the speed drivers were travelling. On 92 out of 96 trials, drivers took the bypass ramp; on 2 out of 96 trials the same driver went on to the roundabout and took the exit to Hwy 406 N there. Two trials involved errors in that drivers took the roundabout exit to Welland, not the correct destination. This was due to forgetting in both cases. In summary, the positive guidance plan on the westbound approach was also successful in leading almost all drivers heading to Hwy 406 N to take the ramp bypass. A third positive guidance question on the westbound and north to westbound approaches was the effectiveness of the positive guidance plan in getting drivers heading to Welland into the left lane well before entering the tunnel. This was tested in Drives 1b (Regional Road 27 WB to Welland and 2b (Farr Road NB to Welland). There were a total of 96 trials for drivers heading to Welland. Again a stalled vehicle was used to force drivers initially into the wrong lane (the right lane) for their destination, so that they would have to change lanes. In Drive 1b the majority of drivers initiated their lane change to the left lane at 750 m from the tunnel, suggesting they were using the sign information. In Drive 2b all drivers initiated the change to the left lane in the intersection. Since this was the lane that drivers are supposed to turn into, and since drivers could not have read the sign from this point, it does not provide any comment on the quality of the sign information. In all but 1 out of 96 trials drivers were fully in the correct lane at least 350m before the tunnel. The remaining driver completed the lane change inside the tunnel. The driver who did this had driven through the tunnel correctly on two previous occasions. Thus the plan was successful in leading almost all westbound drivers to Welland to choose the correct lane into the tunnel. The positive guidance plan on the approaches to the roundabout itself appears to have been somewhat less successful. All drivers exited the tunnel in the correct lane (left) for their destination. But by the time they entered the roundabout however, about one in five had moved to the right lane, an unnecessary lane change, suggesting uncertainty about the correct lane. On Drive 1b, drivers slowed steadily on exiting the tunnel, from 84 km/h in the tunnel to 20 km/h (slowest) at 25 m from the roundabout entry. On Drive 2b, behaviour was similar with speeds being reduced from 80 km/h to 20 km/h. Interestingly, on both Drives 1b and 2b, drivers also slowed after the tunnel, when the ramp bypass lane and sign became visible, by 6 9 km/h, even though this was not their destination. The amount of slowing was less than the slowing evident on Drives 1a and 2a (16 km/h), for which the ramp bypass was the destination. With respect to behaviour while in the roundabout, on Drive 1b and on Drive 2b, about one third made an unnecessary lane change. With the current three leg roundabout configuration, either entry lane allowed drivers to reach their destination. UNIVERSITY OF GUELPH Page 67

102 The unnecessary lane changes both on the approach to the roundabout and while in the roundabout suggest uncertainty about which lane was correct. As it happens both lanes were correct, nonetheless unnecessary lane changing is undesirable Eastbound Approaches to Hwy 406 N (Drives 3 and 4) A fourth positive guidance question was the effectiveness of the positive guidance plan in getting eastbound drivers heading to Highway 406 N into the left lane and taking the second exit to the right. On the eastbound approach to the roundabout (Drive 3), drivers encountered a ground mounted map sign followed very shortly by an overhead guide sign. See Figure 36. Of 48 drivers heading for Hwy 406N, 37 entered correctly in lane 1 or 2, but a substantial number (23%) entered the roundabout incorrectly in lane 3. A second eastbound approach involved initially going northbound on Wellington St., turning right (eastbound) onto Regional Road 27 and then taking the roundabout to Hwy 406 N (Drive 4). This drive allowed less lead time, as compared to Drive 3, to determine which was the correct lane to enter the roundabout. Again a substantial number of drivers entered incorrectly (31%) in lane 3. While in the roundabout on Drive 3, 40%, and on Drive 4, 29%, made an unnecessary lane change. These unnecessary lane changes reflect uncertainly about which lane to be in. These findings suggest problems with the positive guidance plan for eastbound drivers. UNIVERSITY OF GUELPH Page 68

103 Figure 36: Guide signs encountered on eastbound approach Southbound Approach (Drive 5) A fifth positive guidance question was the effectiveness of the positive guidance plan in getting southbound drivers heading to Port Colborne into either entry lane and taking the second exit to the right. On the southbound approach to the roundabout (Drive 5), drivers were given a destination of Regional Road 27 eastbound. Drivers encountered two ground mounted map signs followed very shortly by an overhead lane designation sign. There is a two lane approach with a third added as a slip lane to Welland in the last 135 metres. See Figure 37. Figure 37: Guide signs on southbound approach With the current three leg roundabout, either entry lane allowed drivers to reach their destination. Of 48 drivers, 3 (6%) forgot their destination and incorrectly took the slip lane to Welland and one driver took the roundabout backwards. UNIVERSITY OF GUELPH Page 69