Microcomputer-based Data Logging Device for Accelerometry in wimming. Ohgi Faculty of nvironmental Information, Keio Univ., Japan H. Ichikawa Doctoral Program of Health and port ciences, Univ. of Tsukuba, Japan ABTRACT : The authors developed a microcomputer-based data logging device to evaluate the swimming stroke motion. This newly developed device enabled observation of the tri-axial wrist accelerations in swimming over a one hour-period with no restrictions. In this study, we looked at the change in wrist acceleration affected by swimmer's fatigue during intensive interval training. The direct linear transformation method was used for kinematical quantitative evaluation. A blood lactate test was used as a physical quantitative evaluation of the swimmer's fatigue. It was observed that swimmers' stroke motion changed with increased exhaustion. And these changes were also observed in acceleration patterns. INTRODUCTION We had previously reported that there are some common characteristics with our measurement of tri-axial wrist acceleration during swimming(1999, 00). In order to observe stroke motion by such accelerations, it was necessary to develop long-life non restricted acceleraton device. The primary purpose of this study was to develop a new data logging device that could obtain tri-axial wrist acceleration during swimming. The second purpose of this study was to analyse the change in swimmer's wrist acceleration as exhaustion sets in high intensive interval training. DATA LOGGR PCIFICATION Figure.1 shows the acceleration sensor device developed. The technical specifications of this device are shown in Table1. We attached this logger onto the distal of the swimmer's left forehand tightly using tape and bandage. To start the data logger, we use of a trigger synchronized with the video frame counter. After the experiment, the logger was connected to a PC by paralell cable, and the stored acceleration data downloaded and then browsed immediately through the graphical data browser. 1
Fig.1 Tri-axial acceleration data logger for swimming Table 1 Logger pecifications Dimensions Water Resistent Acceleration ensors Microcomputer ampling Rate Data torage Battery Lifetime 88mm 21mm (0g: including battery) 0m Depth by alminium alloy cylinder ADL2( G) 2 (Analog Devices) PIC(CPU, ROM, RAM, 8ch 12bit A/D, I/O) Up to 128Hz with digital trigger start 32Mbit RAM (2 million data) 1. hour with 128Hz PRIMNT PROTCOL Four male and one female triathlon athletes were involved in the study. All subjects had submitted their informed consent in advance. To determine each subject's swimming velocity, three speed tests (0m 3) were performed before the main experiment. This provided estimated onset of blood lactate accumulation speed (OBLA speed) for each subject and defined a 0m target time(oblat) for the 1st set of the experiment. The experimental protcol is illustrated in Fig.2. The subjects had to swim eight 0m lengths to their own target time, as best they could. The target time of the 1st set was calculated from OBLA speed and called OBLAT. Their target time at 2nd and 3rd set were 3 and 6 seconds faster than that of 1st set, called OBLAT 3 and OBLAT 6 respectively, which should lead to exhaustion after the 3rd set. The interval was 1 seconds between each and heart rate was measured immediately after each set. Blood lactate was sampled 3 minutes after each set. A Pro(Model LT-17, Arkray Inc.,) was used for the blood lacate sampling test. 2
HR HR HR HR 1st et Rest 2nd et Rest 3rd et 1min Fig.2 Protcol of xperiment RMOT CONTROLD UNDRWATR CAMRA A remote controled underwater pan-tilt camera was also developed for kinematic observation of the swim stroke. A remote controled pan-tilt camera(vi-d, ony Inc.,) was waterproofed in an acryllic case(fig.3) able to stand with a depth of m. Two underwater cameras were controled simultaneously using R232C control command on the poolside. The control software configures the camera's pan-tilt angle, rotation speed, zooming, focusing, shutter timing and exposure settings. One underwater camera was set at the bottom of the pool diagonally to the left front of the swimmer, while the other was set diagonally to the left behind the swimmer. A frame counter, synchoronized with the data logger, overlaid a time code on the video frame. The underwater swimming motion was recorded on digital video. For three dimensional motion analysis, the calibration markers were set beneath the lane line as shown in Fig.3. The subjects were marked on their shoulder, elbow, wrist and metacarpophalangeal joints. A manual digitizing process was then used for these joint coordinates. The digitized joint coordinates were reconstructed with respect to the global coordinate system by the direct linear transformation method. wimmer's Traveling Direction 2.0m 2.0m Camera1 1.2m Camera2 11.7m 12.m 13.3m PC Monitors and Recorders Video Frame Counter Fig.3 Remote controled underwater cameras 3
RULT FATIGU TIMATION Figure.4 shows the average swimming velocities and blood lactates of the five subjects. The subects swam almost to their target time in each set. It was observed at the end of the final set that they were near exhaustion, where their blood lactates were between 6.8-14.9mmol/l. In particular, sub.c, D and finished with a blood lactate over 11mmol/l. The intensity of interval training was enough to exhaust them. Velocity[m/s] 1.6 1.4 1.2 A C D 1 2 3 et B Blood [mmol/l] 16 14 12 8 6 4 2 Pre 1 2 3 et C D B A Fig. 4 Change of average swimming velocity and blood lactate ub.a ub.b ub.c ub.d ub. 0.2 0. 0.7 1 1.2 1. 0. 1 1. 2 0.2 0.4 0.6 0.8 1 1.2 1.4 0.2 0. 0.7 1 1.2 1. 0.2 0. 0.7 1 1.2 1. 1.7 - - - - - - - - - 0 0 0 - - 0. 1 1. 2 0.2 0. 0.7 1 1.2 1. 0.2 0.4 0.6 0.8 1 1.2 1.4 0.2 0. 0.7 1 1.2 1. 0.2 0. 0.7 1 1.2 1. 1.7 - - - - - 1 1 1 1 1 0.2 0.4 0.6 0.8 1 1.2 1.4 0.2 0. 0.7 1 1.2 1. 0. 1 1. 2 0.2 0. 0.7 1 1.2 1. 0.2 0. 0.7 1 1.2 1. 1.7 - - - - - - - - - - -1-1 -1-1 -1 Time[s] Time[s] Time[s] Time[s] Time[s] Fig., and -axis accelerations at 8th trial in 3rd set ACCLRATION ince the and -axis of the sensor was parallel to the palm plane, both showed a significant peak acceleration at the moment of entry. -axis acceleration always showed a negative acceleration at this time. Therefore, from the negative peak acceleration, we could determine each entry timing. We could then extract, and -axis acceleration for each stroke. Figure. shows all subjects on the last trial(8th) of the 3rd set. 4
DICUION There are two characteristics among swimmer's acceleration patterns. On one hand, certain characteristics are common between subjects. For example, the impact acceleration on entry gives rise to an and -axis peak. On the other hand, there are subjects-specific characteristics. For example, the timing of the local maximum or minimum of every axial acceleration is different fom each subject. In addition, the overall acceleration patterns remained even though swimming velocity changed. In this section, we will focus on the results of sub.c and discuss this particular stroke motion. ub.c showed 13.2mmol/l blood lactate after the 3rd set and indicated complete exhaustion. The 8th trial acceleration on the 3rd set was clearly different from that of the 1st trial in the same set which was before the exhaustion. Figure.6 shows the tri-axial wrist acceleration on 1st and 8th trials in the 3rd set. The bottom graphs show elbow angle and angular velocities during the underwater phase. The reason that duration of elbow angle and angular velocity graphs are shorter than that of acceleration, is that we are illustrating only when the shoulder, elbow and wrist joint could be digitized. Figure.7 illustrates the stroke paths on 1st and 8th trial in the 3rd set. The shoulder(), elbow() and wrist(w) coordinates are plotted. We can see that both and -axis' local maximum at the middle of stroke on the 8th trial was lower than that of the 1st trial's. Accoding to the time axis of these figures, the stroke duration on the 8th trial was shorter than that of the 1st trial. This indicates that sub.c accelerated his stroke rate on exhaustion. Ohgi, et al. (1999, 00) described that the local maximum of the -axis acceleration at the middle of the single stroke corrensponds to the beginning of the insweep motion in front crawl. They suggested that a swimmer makes the insweep motion with the hand plane having an angle of attack. Decreasing of the -axis local maximum would be caused by the change of the angle of attacks between palm plane and its velocity vector. From Fig.7, after moment of entry, the subject's left wrist moved outward and downward on the 1st trial. Then it moved inward to his trunk. In contrast, on the 8th trial, the wrist moved straight forward and downward firstly, then moved backward. Thus, the insweep motion had disappeared on the 8th trial. We suggest that the backward stroke motion caused decreasing of the local maximum of the -axis acceleration. Ohgi, et al. (1999) reported that when -axis acceleration decreases to 0ms -2, this corresponds to the stretch phase. On the 1st trial, the -axis acceleration at 0 ms -2 was 0.37sec. However, on the 8th trial, the duration steeply decreased and reached 0 ms -2 after entry. Then acceleration increased without steady state. Therefore, the shortening of the stretch phase caused the total stroke time to decrease. vidence of stroke duration shortening can be seen in the stroke paths of Fig.7. troke length per stroke was remarkably shortened under exhaustion. Among the three joint coordinates, the wrist joint on the 1st trial moved forward after entry. However, on the 8th trial, it moved downward following entry, namely the negative direction of the -axis with respect to the global coordinate system. After entry, the swimmer started a downsweep motion in the fatigue situation without a stretch motion. A small local maximum() also appeared after the global maximum of -axis acceleration(). We see that the -axis local maximum on the 8th trial decreased below that of the 1st trial. The finish of the insweep motion which is equal to the beginning of the upsweep motion, corresponds to the time at. In Fig.7, the time at is also explained. The two bottom graphs in Fig.6 show the change of elbow angle(θ e ) and its angular velocity(ω e ). From these angular parameters, almost
corresponds to the point of maximum elbow flexion. Before exhaustion, this occured at 0.93sec. This then shortened to 0.83sec at the fatigue situation in the 8th trial. Usually, a swimmer flexes his elbow joint at the end of insweep, that coincides with the begining of upsweep. Local maximum is probably caused by a centrifugal acceleration component made by the elbow extension at the beginning of the upsweep motion. ub.c released the left arm and hand with an elbow angle at about 0[deg] at the 8th trial(fig.6). This also explains that the centrifugal acceleration made by elbow extension save rise to the -axis local maximum at. o far, in order to observe the underwater stroke motion, we had to prepare underwater cameras. Although we used them, it was difficult for us to determine which stroke phases were changed during daily training or at a swimmer's fatigue point. The data logger device helped us to observe a swimmer's stroke pattern relatively easily. From that we could obtain several key characeteristics of the uderwater stroke motion using an acceleration sensing device. - - - 0 0 - - 3rd et 1st 0.2 0. 0.7 1.0 1 11. 0.2 0. 0.7 1.0 1 1.2 1. Time[s] 0.2 0. 0.7 1.0 1.2 1. - - - 0 - - 3rd et 8th 0.2 0.6 0.8 1.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Time[s] 2 Angle[deg] 180 160 1 1 0 0.2 0. 0.7 1.0 1.2 10 800 0 0 1. -0 Angular Velocity[deg/s] Angle[deg] 180 160 1 1 0 e e 0.2 0. 0.7 1.0 1.2 10 800 0 0-0 Angular Velocity[deg/s] Fig.6 Accelerations, elbow angle and elbow angular velocity during high intensive interval training 6
11. 12 12. 13 13. W -0.2-0.6 0.0-0.2 11. 12 12. 13 13. -0.2 W -0.6 0.0-0.2 W -0.6 W -0.6 1.2 1.0 0.8 0.6 1.4 1.2 1.0 0.8 Fig.7 Underwater stroke paths of swimmer's shoulder(), elbow() and wrist(w) joints CONCLUION Tri-axial acceleration during swimming training data was obtained for over an hour using a microcomputer-based acceleration sensor device. Five triathletes swam intensive interval training sets with this device mounted on their left wrist. Underwater stroke motion and acceleration patterns changed under the fatigue situation. Using this newly developed device, we are able to obtain the characteristics of the swimmer's stroke motion during swimming training both easily and continuously. RFRNC Ohgi,., Ichikawa, H., Miyaji, C. (1999) Characteristics of the forearm acceleration in swimming, Biomechanics and Medicine in wimming VIII, Keskinen, K.L. et al., ds, pp77-82. Ohgi,., asumura, M., Ichikawa, H., Miyaji, C. (00) Analysis of stroke technique using acceleration sensor IC in freestyle swimming, The ngineering of PORT, A.J. ubic and.j. Haake ds., pp03-11. 7