Tuesday, March 21, 2006

Eldercare Robotics: A Personal Mobility Aid using shared control systems

U0205119 - Ek Li Ling Some elders may suffer from conditions such as macular degeneration and cataracts, which result in vision impairment. Others may find it difficult to move around due to loss of motor coordination. There are many types of existing aids available to help elders maintain their independence and mobility, but they provide limited help. Canes and dogs help guide the way, but they do not provide a stable platform to support the elderly in the event of a trip, while although walkers and scooters may have stable platforms, they do not solve the problem of vision impairment. A walker has been designed by the Medical Automation Research Centre (MARC) at the University of Virginia which is aimed towards helping elders who are both vision and motion-impaired. This three-wheel walker uses sensor technology to detect obstacles and also a navigation system which controls the steering well when collisions with obstacles need to be avoided. The walker is an example of a passive robot. Passive robots are robots that can steer their joints, but require a human to move them. The walker is a passive robot because it can only change the direction of its front wheel and cannot move forward on its own. Such a feature was decided because users need to feel that they are in control instead of being led by, or having to chase after the walker. The walker also features rear-wheel collision detection sensors that will notify the user when the back wheel may get caught by an obstacle. Three-wheel walkers have a wide wheelbase, and commonly face such a problem, especially through doorways. The walker aims to be an intelligent shared control system, which can be seen as a combination of two control systems that are inputting control signals to the walker frame. One system is the human, providing the moving force and steering the walker towards their destination. The second control system is the steering by the walker to avoid obstacles and prevent falls. The intelligent walker allows the users varying degrees of control, according to the user's abilities and needs. This varies from complete control to collaborative control with the control system. Such a feature is aimed at allowing the users to feel in control, while at the same time increasing the ease and safety of their travel. It seems rather dangerous for elders to have a walking aid which can change direction on its own. Therefore, it is very important that the control agent does not result in big movements unexpected by the user. For example, if the walker’s movement differs significantly from the user’s expectation, the walker might cause the user to fall, instead of preventing a fall! In addition, users expect that the walker will maneuver in the way that they push it. The control system must not make the user feel as if the walker is unresponsive or non-obedient. Therefore, an important design issue is the ability to integrate the two control systems to provide a smooth sense of shared control. Deciding where and when to vary the degree of autonomy given to the control agent is an active area of research. The navigation system will automatically vary the level of intervention according to the user’s wishes and the prescence of obstacles in the environment. The control agent has to be submissive, but alert. Currently, when no danger or difficulty is detected in the environment ,and the user and the control agent express different desires on how to steer the walker frame, the control-agent must give in and the walker will give full control to the user. When the walker attempts to steer around objects or between spaces, the walker is controlled by a mix of user and control agent commands. Finally, the walker control agent only takes full control when a collision or drop-off is imminent. This walker is still in the development stages. Research is being done on different control systems to improve the user’s experience of this product. The success of this product depends largely on how well the control system reads the user's intentions and collaborates with the user. Hopefully, this product can truly extend the mobility and independence of our elders. References Medical Automation Research Centre, Projects, Eldercare Robotics: A personal Mobility Aid. http://www.medicalrobotics.com/projects_eldercarerob.html Wasson, G., Gunderson, J., Graves, S. and Felder, R. 2001. Effect Shared Control in Cooperative Mobility Aids. FLAIRS ’01: 509-513. http://www.medicalrobotics.com/pdfs/projects/mobility_aids.pdf

3 comments:

Anonymous said...

U0204511 Tan Chin Hiong

This walker seemed like it needs a powerful image recognition algorithm to distinguish surrounding dangers (eg. oncoming cars). Also, with its main sensors all mounted so low, it will not be able to detect tall collision objects like sign post, tall branches, flying soccer ball approaching. So this may be an area for improvement.

Another thing is that the metal frame seems quite big and wide. So maybe there will be problems going through narrow passages. Perhaps the body could be streamlined for better mobility without compromising stability.

My 2 cents.

u025445b lau duan neng said...

This product would certainly help the elderly in moving around as well as those who have difficulties moving around!

I do agree with Chin Hiong that the robot needs a powerful image recognition algorithm to distinguish surrounding dangers and that improvements are needed as all the main sensors are mounted so low, which resulted it not being able to detect tall collision objects. Perhaps a speaker can be added in as well so that it can inform the user of any possible danger present around him. But this would mean that additional circuitries are needed to distinguish the object and then transfer the information to the user in real-time.

Medical said...

U0205119 Ek Li Ling
Currently, the walker does not rely on any image recognition algorithm at all, but relies one sensor technology to detect obstructions close to it. For example, the sensors are similar to those equiped on cars to help with parking. They do not identify what the obstructions are, but are able to detect the proximity of the obstruction to the car/walker.

I suppose this sensor device will be more simple and economical to suit the needs of the elders who only have slightly poorer eyesight, as they should still be able to see sign post and tall branches. However, maybe a more advanced model, and definitely more expensive model with image recognition funtions can be implemented for those who need the extra help.

For the metal frame, I agree with you that it is very bulky. Hopefully, they will come up with a better structure.
As for the sound sensor, there are considerations to implement this feature as well.Instead of a sound sensor which may distrub people around, a vibration warning system is also being considered. I guess both options have their pros and cons.

Thank you for your comments!