US researchers are developing a wheelchair that uses a robot to teach young users how to use it safely at their own pace, in the hope that it will lower the cost and improve accessibility to wheelchair training for children with a disability.

You can read how Dr Laura Marchal-Crespo and colleagues at the University of California at Irvine developed and tested the robotic wheelchair with a group of children without disabilities and one child with cerebral palsy, in the 13 August issue of the open access Journal of NeuroEngineering and Rehabilitation.

The researchers developed ROLY, RObot-assisted Learning for Young drivers, because the current way of training young children to use a powered wheelchair requires a trained therapist to be with them nearly all the time, mostly using the hand over hand method. This is “expensive and labor-intensive” said Marchal-Crespo in a prepared statement.

So to reduce the cost and make the training more accessible, they have “have developed a robotic powered wheelchair system on which young children with a disability can safely develop driving skills at their own pace with minimum assistance,” she added.

At the centre of ROLY is the idea of “haptic guidance” where the robotic wheelchair steers itself along a course marked by a line on the floor using computer vision, and this “knowledge” is used to decide how much control to apply to the joystick that the young learner controls.

In this sense it is like having the therapist with the hand over the learner’s hand and gradually surrendering control to the learner as they improve. ROLY applies just enough force, in addition to that applied by the learner, to lead the wheelchair in the right direction. This force also changes in line with the varying performance level of the learner.

The goal for the learner is that they have to catch a little mobile toy robot that runs along in front of the wheelchair, as if in a game of “robot tag”.

When the child catches the toy robot it does a little dance and the wheelchair plays a little tune.

In their paper the researchers describe the design of ROLY’s computer vision and control system. They also show data from the pilot study where ROLY “taught” 22 children without disability how to use the wheelchair in a single session. The point was to establish normative values of learning rates, a sort of baseline from which to start using ROLY with children of varying disability.

The results showed that with ROLY, ie with haptic guidance, the children without disability learned to steer significantly better than when trained without guidance.

In a further case study, they used ROLY to train an 8-year old child with cerebral palsy whose motor ability was severely impaired. The study followed the same single-session protocol as that used with the non-disabled children. The researchers found that:

“This child also improved steering ability after training with guidance from the joystick by an amount even greater than the children without motor impairment.”

They concluded that ROLY not only provided a safe and enjoyable context for automating the training of wheelchair drivers, it also:

” … enhanced motor learning by the non-impaired motor system, presumably by demonstrating through intuitive movement and force of the joystick itself exemplary control to follow the course.”

They also concluded that the case study showed that haptic guidance was able to help a child with cerebral palsy and an impaired motor system to benefit from driver’s training.

Marchal-Crespo said they could see a way to create a training experience “that compares favorably with the fun children experience with the best amusement park rides, but that facilitates the development of driving skill”.

A robotic wheelchair trainer: design overview and a feasibility study.
Laura Marchal-Crespo, Jan Furumasu, David J Reinkensmeyer
Journal of NeuroEngineering and Rehabilitation 2010, 7:40 (13 August 2010)
DOI:10.1186/1743-0003-7-40

Additional source: BioMed Central.

Written by: Catharine Paddock, PhD