NOVEL IMPLEMENTATION OF AN ADAPTIVE CONTROL BRAKING SYSTEM FOR A WHEELED WALKING AID

Size: px

Start display at page:

Download "NOVEL IMPLEMENTATION OF AN ADAPTIVE CONTROL BRAKING SYSTEM FOR A WHEELED WALKING AID"

Adele Russell
5 years ago
Views:

1 NOVEL IMPLEMENTATION OF AN ADAPTIVE CONTROL BRAKING SYSTEM FOR A WHEELED WALKING AID E. Young*, E. Coyle, K. Sullivan, A. Toner Dublin Institute of Technology, Kevin Street, Dublin 8, Ireland. Eileen.young@dit.ie Keywords: Walking Aid, Adaptive Control Braking, Rollator Abstract A wheeled walking aid with an adaptive braking system is described. The frame of the prototype is based on a combination of features of standard available wheeled walking aids. A braking scheme has been designed using hydraulic disc brakes to facilitate accurate sensitive controlled stopping of the walker by the user, and if called upon by automatic action. Braking force is modulated via a linear actuating stepping motor. A microcontroller is used for control of both stepper movement and supervisory control. An encoder is used to supervise walker movement in terms of time, distance and speed. The focus of the research is that of improving upon ease of movement and safety enhancement to people requiring mobile aid assistance. 1 Introduction The primary motivation behind this research was to consider mobility aids available on the market and to improve upon their performance and usefulness, with focus on stability, ease of control and safety. In particular, the intention was to enhance the prospects of people with conditions such as muscular dystrophy, by providing them with the mobility necessary to maintain a reasonably independent lifestyle. The inspiration was that of a colleague, who in spite of a having developed such a condition has maintained his position in the work force and who continues to commute on a daily basis into the city of Dublin. Whilst both electric and manual wheelchairs serve a useful purpose, the ability of the user with disability to walk with assistance over short distances, is highly desirable if at all possible, and helps to boost individual morale and assists in maintaining muscle tone. 1.1 Engineering Terminology as Applied to Disability To gain some perspective on engineering in terms of people with disability it is essential to carefully select the terminology to be used. Terminology sets the basic tone of any research and if used correctly can make the world of rehabilitation a more inclusive environment, alternatively if used carelessly terminology can be the basis of further discrimination. Definitions of disability used can influence the way in which non-disabled people respond to disabled people. Chadwick in his work concludes that a definition can place limits not merely on what is possible, but what is thinkable in an organisation [5]. The World Health Organisation (WHO) introduced its text International Classification of Impairments, Disabilities and Handicaps (ICIDH) in 1980 [22]. This text proposed to develop a global common language in the field of disability. It advocates describing people as having a disability rather than being disabled. The ICIDH has played an important role in the development of policy in relation to people with disabilities throughout the world. A fundamental tenet of the European Union has been the development of a disability strategy based on the United Nations Standard Rules on the Equalisation of Opportunities for Persons with Disabilities [18] in association with the International Bill of Human Rights. This has been further developed by implementation of a European Action Plan in 2003 [7]. 1.2 Disability, Rehabilitation and Mobility The estimated number of people who require rehabilitation services at any one time is 1.5% of the world population, i.e. about 90 million people. The number of people with disability is estimated at 7% to 10% of the of the population [19]. The group representing the aged population will have an increasing impact on disability numbers in the most countries. Although the age of 65 is used as determining of old age, it has been noticed that actual old age and related needs of services starts around years of age. At age 75+, 72% of people have Functional Limitation of which 41% have Severe Functional Limitation [11]. 2 Commercially Available Walking Aids Walking (Zimmer) frames are lightweight units, which come in a range of styles. They can be fixed height or adjustable, with fixed or folding frame. Two-wheeled walkers are similar in design to Zimmer frames, with two wheels positioned to the front of the frame. Both walking frames and two-wheeled walkers are pick-up walkers, i.e. they must be picked up and moved to a new location to take a step. They have advantages in that they are both stable in design and cheap to purchase. They also have the disadvantage of the user having to stand unaided upon lifting the frame, and their use can encourage an unnatural gait.

2 Another category of wheeled walker is that of the threewheeled walker, designed with two fixed rear wheels and a swivel front wheel. Advantageous features of these units include foldable design and a selection of handles and/or arm-troughs to suit user preference. They may be fitted with either spring-loaded or loop brakes. A major disadvantage is that they are inherently unstable. Four-wheeled walking frames are also available, designed with two front swivels and two rear fixed wheels. These units are foldable and have a basket, which may also be used as a resting seat when stationary. Again, they may be fitted with either spring-loaded or loop brakes. 3 Body Kinematics A number of reports have compared basic gait factors, including stride length, cadence, and velocity, and have gauged user satisfaction with different walker designs [13,14,17]. Studies have also been conducted into the energy required by the user using different frame types and usage pattern [3,9,10]. Studies have focused on different areas of the body as in the upper extremities, the lower extremities, the load placed on the walker or the load placed at the handles of the walker. However, in most cases these studies do not give a complete quantifiable picture of the dynamics of using a walker [6,15,16]. Adrezin et al instrumented a standard walker, and parameters measured by the team in carrying out assessments have included those of axial forces, and anterior-posterior and medial-lateral bending loads at each of the four legs [1]. It was found that the wide range of occurrence of user disabilities led to differing walker usage patterns. Hence it was concluded that a need exists for future research leading to better understanding of walkers and new walker designs, which will address the varying needs of users. Power assisted walkers have been designed to support people with weak lower limbs. The use of walkers is difficult for hemiplegic users. The difficulty is that the weakening of one side due to their illness makes it difficulty to control the direction of the walker. Egawa et al have developed a powerassisted walking support system to compensate for this difficulty [8]. The power-assisted walker allows hemiplegics better control of their walk by compensating for force input bias caused by imbalanced gait. Intelligent robotic walkers are being developed for a range of different purposes. These walkers can be either driven or passive but the common ground is in their aim to enhance user mobility and safety. Instrumentation is used to monitor user intent and to compare with information in the environment to determine the walkers movement [12,20,21]. The outcome suggests that a wider range of user needs are being met. 4 Users Requirements Development of the system in the current research has mainly focused on the physical requirements of and the knowledge imparted by a colleague, for whom this prototype has been developed. His condition is such that he cannot stand independently and must be supported at all times. At present, the only viable option available suitable to his requirements is that of the use of crutches. The approach is to place one crutch forward and then swing the opposite foot forward. When stable he then switches over and moves the alternate crutch forward and then the other leg. At all times his centre of gravity is maintained to the front of his feet position, this locks his ankles and knees and enables his legs to weight bear. This is a very physically demanding process, is wearing on shoulders, elbows and wrists, and is becoming increasingly difficult with progression of time. As our user s centre of gravity is to the front of his feet, there is always a forward force on any walking aid he uses. This forward force increases the requirements of the walker in the areas of safety and braking. Standard walking frames or two-wheeled walkers are not a suitable option, as both of these require the user to stand unaided and to lift the unit forward before taking a step. This has led to the option of developing an alternative wheeledwalker, which would overcome the difficulties associated with existing available models. 5 Walker Feature Design A design decision was made to utilise the base of a commercially available four-wheeled rollator, Voyager by Worldwide Mobility, for stability. This base has two swivel wheels positioned at the front and two fixed wheels at the rear. The swivel wheels facilitate manoeuvrability while the fixed wheels encourage the unit to move in a straight line. The combination of two fixed wheels and two swivels and a large footprint gives the unit both controllability and stability, making for a good general-purpose walker. To give our user upper body support and to allow for loss of power, the handles have been changed to arm rests. The arm rests are height and position adjustable for maximum comfort. The prototype walker is shown in figure 2. The brakes have been changed to hydraulic disk brakes. Brake disks have been attached to both the rear wheels with their callipers attached to the frame. Disk brakes, as they are not in contact with the ground, have a high accuracy and repeatability. The brakes use a closed hydraulic system, which is very responsive. Hydraulic disk brakes have been specifically chosen to give a controlled stop and for the ease of activating two disk brakes at the same time. An area of concern is that due to the slow speed of a walker, the brakes will be grabby and will stop the walker instantly once the calliper comes in contact with the disk surface. To overcome this potential problem the brake pressure can be modulated to give gradual application.

User PID Fm Supervisor y Controller P bset + Fuzzy Logic PWM Stepping Motor + Drive Hydraulic System Pb µπ ( 9 3 r 3 1 r2 ) r w Calliper Fb Walker + User Dynamics S w Stored Ste p Profiles Hydraulic

3 User PID Fm Supervisor y Controller P bset + Fuzzy Logic PWM Stepping Motor + Drive Hydraulic System Pb µπ ( 9 3 r 3 1 r2 ) r w Calliper Fb Walker + User Dynamics S w Stored Ste p Profiles Hydraulic Pressure Encoder Figure 1: Shared Control Walker Prototype Control Loop. Linear Actuator Pressure Disk Brakes Encoder Battery Figure 2: Prototype Wheeled Walker. Brake Lever Brake Switch Stepping Motor Controller Microprocessor 68HC11 Initially for testing purposes the brakes will be activated by a lever. Once the fundamental braking information has been recorded and quantified the brakes will be changed to automatic controlled operation. The automatic brakes will have an emergency / parking brake switch and an activate brake in normal walking switch. The ideal situation will be that the supervisory controller will determine brake application and not require user intervention. The control loop of the prototype system is shown in Figure 1. Activation of the brakes is via a stepping motor controlled by an on-board electronic drive. The stepping motor is a linear actuator and is a high-speed high torque unit (to facilitate modulation). The motor is attached to the brake cable and will activate a hydraulic brake piston, as required. A pressure transmitter in the hydraulic line gives feedback of the pressure being applied to the brake and the work being done. Information on distance, speed and acceleration are measured by using a differential encoder attached to the rear wheel of the walker. The two pulse trains from the encoder are fed into the microcontroller, which calculates magnitude variation of walker movement. Instructions will be delivered to the motor drive from the microcontroller. The microcontroller will also collect and process the encoder data, pressure transmitter and switch information. The microcontroller has two functions; it serves as both the process controller and the supervisory controller working in parallel with the user (Figure 1). As the process controller it controls brake position to a selected control strategy. As the supervisory controller, the microcontroller uses historical data as well as current data to determine whether and to what degree the brake should be applied. Decisions are based on stored profiles and stored maximum values and the purpose is to ensure user comfort and safety. 5.1 Control Philosophy The encoder on the walker wheel measures distance travelled and the microcontroller calculates speed and acceleration. This data is the basis for profiles of the walker movement and will be measured and recorded over a number of steps for different users. A correlation between profiles for steps for one user and then for different users will be made. A simplified velocity step profile from initial experiments is displayed in Figure 3. The profile consists of an acceleration phase and a deceleration phase. It was used to determine component size when building the walker.

4 Velocity m/s Walker Velocity Profile Time s Figure 3: Simplified Velocity Profile from Experimentation Velocity m/s Walker Expected Velocity Profile Time s Figure 4: Expected Velocity Profile Type from Prototype The walker prototype will be used to develop step profiles. These will be analysed and broken into a defined number of stages depending on what their natural profiles are. Therefore if the natural profile comprises an acceleration phase, a constant speed phase and finally a deceleration phase, as shown in Figure 4, the movement of the walker will be reduced to three stages. This step profile becomes the controller setpoint. At each stage, maximum distance, speed and accelerations for the safety of the user will be achieved from experimentation. Thence, as a user takes a step, should the parameters be outside the maximum values, the walker brakes will automatically come into action. The profile under design will also have an adaptive learning facility. This will be utilised to take into account and make corrections for a user who may be tiring. It will also be utilised to accommodate a change of user. Over a period of typically 10 steps the parameters of each step will be measured and correlated. This will then become the normal step. Around this step there will be bands of maximum distance, speed and acceleration, which the walker controller will obey. Therefore, if the user is taking a normal step and the walker is following the accepted norm, the brakes may not need to operate. However, if the walker is not following the norm the brakes will come into action, initially gentling, but as the difference increases, engaging with more effect. There will be a dead-band around the norm where no intervention will take place. The normal step will take in new data at each step and update the profile. The adaptive step profile will be the setpoint for the control loop (Figure 1). Initially the control loop will be a PID loop to test the system. If the system is grabby, pulse width modulation will be used to improve the braking characteristic. Consideration is also being given to use of fuzzy logic as an alternative control strategy. As there are a number of step types made when using a walker, the supervisory controller will determine which step type is currently being taken. The step types include: full walking speed step forward, small step forward, small step backward, and turning movement. Each of these movements will have a separate profile, which the walker will automatically control to. The type of step will be determined by the initial movement of the walker, the historical movement of the walker and the work the brake is doing. The intention of the user and the decision of the supervisor as to the intent of the user may disagree at times. In such instances the movement of the walker and the pressure feedback will be constantly monitored. In the event that the brake is beginning to activate and in reality the user is still driving the walking aid forward, there will be more work at the hydraulic pressure than expected. This information is used to update the step type selected. 6 Discussion The development of this walking aid has been inspired based on the mobility needs of an individual user. Upon completion of the overall system software and hardware design, and following upon an extensive testing programme, the success of the proposed prototype in meeting user requirements and expectations will be determined. Modifications and improvements will be made as required. It is hoped the walking aid will have a broader appeal than to a single user. It is envisaged that the unit will be of aid to a wide range of people with walking difficulties, especially whilst ambulating on sloped ground, and to users with an inherent tendency to loose balance and control owing to a clinical condition An important area of interest is to make the transfer from user braking to automatic braking smooth so as to give the user a feeling of being in control. This area has been researched by Wasson at Al, with development of a walking aid with a controlled front wheel used for obstacle avoidance [20]. This walker is user driven. An important issue highlighted by Wasson is the determination of user intent. User intent and supervisory controller decision must be constantly compared and the supervisory controller intention updated if the paths begin to diverge. A major area of concern with mobility aids is the number of rejections by clients of the chosen aid. The core problem, which leads to rejection, is that of the unit not meeting the user requirements. A recommendation from studies in this area is increased individualisation of walking aids to meet user needs [2,4].

5 References [1] Adrezin R., Cordaro M., Wang F., Fast A. Instrumentation and Computer Interfacing of a Standard Walker to Study User-walker Interaction Dynamics, Advances in Bioengineering ASME, Vol.22, pp ,(1992). [2] Alakarppa I., A Study into the Acceptability of Assistive Devices for Mobility, Technology and Aging: quality of life, ICTA, Sept 12-14, (2001). [3] Baruch I.M., Mossberg K.A, Heart-rate Response of Elderly Women to Nonweight Bearing Ambulation with a Walker. Physical Therapy, Vol.63, pp , (1983). [4] Brandt A., Iwarsson S., Stahl A., Satisfaction with Rollators Among Community Living Users: a follow up study. Assistive Technology, Vol. 25, No.7, pp , (2003). [5] Chadwick, A., Defining Impairment and Disability. Northern Officer Group (NOG), (1998). [6] DeatheA., Pardo R., Winter D. Stability of walking frames. Journal of Rehabilitation Research and Development, Vol.33, No.1, pp , (1996). [7] EU, Equal Opportunities for People with Disabilities: A European Action Plan. COM (2003)650 final, 30 October (2003). [8] Egawa S., Nemoto Y., Fujie M.G., Koseki A., Hattori S., Ishii T., Power-Assisted Walking Support System With Imbalance Compensation Control for Hemiplegics. Proceedings Joint BMES/EMBS Conference Serving Humanity, Advancing Technology Oct 13-16, (1999). [9] Hamzeh M., Bowker P., Sayegy A., The Energy Costs of Ambulation Using Two Types of Walking Frames. Clinical Rehabilitation, Vol.2, pp , (1988). [10] Holder C., Haskvitz E., Weltman A., The Effects of Assistive Devices on the Oxygen Cost, Cardiovascular Stress, and Perception of Nonweight-bearing Ambulation. Journal of Orthopedic Sports Physical Therapy, Vol.18, No.4, pp , [12] Lacey G., Adaptive Shared Control of a Robot Mobility Aid. Proc. Field and Service Robotics (FSR), CMU, August, (1999). [13] Logan L., Byers-Hinkley K., Ciccone C., Anterior Versus Posterior Walkers: A gait analysis study. Developmental Medicine and Child Neurology, Vol.32, pp , (1990). [14] Mahoney J., Euhardy R., Carnes M., A comparison of a two-wheeled walker and a three-wheeled walker in a geriatric population. Journal of the American Geriatrics Society, Vol.40, pp , (1992). [15] Pardo R., Winter D., Deathe A. System for Routine Assessment of Walker-assisted Gait. Clinical Biomechanics, Vol.8, pp , (1993). [16] Simoneau G., Hambrook G., Bachschmidt R., Harris G. Quantifying Upper Extremity Efforts when using a Walking Frame. Pediatric Gait, (2000). [17] Smidt G.L., Mommens M.A., System of Reporting and Comparing Influence of Ambulatory Aids on Gait. Physical Therapy, Vol.60, No.5, pp , (1980). [18] UN. Standard Rules on the Equalization of Opportunities for Persons with Disabilities. United Nations A/RES/48/96, 20 December (1993). [19] WHO, Future Trends and Challenges in Rehabilitation. World Health Organization, (2003). [20] Wasson G., Sheth P., Alwan M., Granata K., Ledoux A., Huang C. User Intent in a Shared Control Framework for Pedestrian Mobility Aids, IEEE/RSJ International Conference on Robots and Systems, (2003). [21] Wasson, G., Gunderson, J., Graves, S., Felder, R. Effective Shared Control in Cooperative Mobility Aids, FLAIRS 01: pp , (2001). [22] WHO, International Classification of Functioning, Disability and Health. Geneva, World Health Organization, (2001). [11] Hypponen H., Disability and Ageing concept definitions. (1997).

C-Brace Orthotronic Mobility System

C-Brace Orthotronic Mobility System You ll always remember your first step Information for practitioners C-Brace Orthotics reinvented Until now, you and your patients with conditions like incomplete spinal