Engineers developing tech for 'terrain smart' wheelchairs

Using research originally developed for the military, engineering experts at Florida State University are creating technology that could give wheelchairs the ability to glide on even the trickiest surfaces -- including icy sidewalks and just-washed floors.

I spoke recently with Emmanuel Collins, director of Florida State's Center for Intelligent Systems, Control and Robotics, about this emerging tech.

Are you, in effect, building all terrain wheelchairs?

We're trying to take existing wheelchairs and make them smarter. We're trying to make them terrain smart. Our current project is focused on how we can improve the control system, so the wheelchairs drive more efficiently on a variety of surfaces. If you're on gravel, we did some experiments to show that you really don't want to turn the wheelchair any more than you have to because you'll lose control. You also don't want to go to fast. You want to adjust the joystick settings -- used to control the wheelchair -- so that it makes it hard for the driver to turn or to have too large a velocity on that surface.

We're adjusting the control system on existing wheelchairs. To be honest, there's a limitation to how terrain smart you can make the vehicle that way. There comes a point where you do have to change the hardware. We think one of the toughest surfaces will be sand. You probably cannot take most existing wheelchairs and make them so they can go on loose, dry sand that's on a beach away from the water. To do that you probably have to change the tires of the vehicle.

We're trying to push the envelope. Think of the control system as the software. How much can we change the software so we have much better performance on a variety of surfaces?

Are you targeting specific surfaces?

We're very interested in slippery surfaces such as icy surfaces, indoor surfaces where the floor has just been washed, something like pavement after it's rained. Even grass is something we're interested in considering. Grass is not super easy to drive on in an electric-powered wheelchair. We're also interested in sloped surfaces. When the users especially go down a slope, sometimes there is a tendency to fall out of the wheelchair. We want to have automatic adjustments to the speed settings, so the user is more stable in the wheelchair seat.

Why are you targeting electric wheelchairs for these improvements?

There are a lot of users. I believe there are over 200,000 users of wheelchairs and it's going to increase. We want to increase their mobility. If you were in a wheelchair, you'd want to be able to go to as many places as you can. Ideally, you'd want to go to the same places that you could go when you were walking. Right now, electric-powered wheelchair users don't go into certain environments. There's a limit on how much they can interact socially. There's a limit on their freedom of movement. We're trying to give them more freedom of movement.

Talk about the technology behind this project.

We were the developers of a lot of the technology. We originally developed it for the military. The military is interested in terrain smart, autonomous ground vehicles. We continue to work in that area. We noticed that a lot of the things we were doing with the autonomous ground vehicles could be transferred to these electric-powered wheelchairs. The problems are not identical, but they're very similar. The primary difference is an autonomous ground vehicle doesn't have a driver. An electric wheelchair has a driver.

How do you use the technology to make wheelchairs terrain smart?

The basic problem is to know how to program the joystick for different surfaces. We take wheelchair users and do testing on various surfaces to figure out the best settings for that surface. We make that testing as scientific as possible. In other words, we're not just looking for qualitative answers. We want quantitative answers. What should be the minimum turn radius on gravel? What should be the maximum speed you travel on gravel? We try to quantify the ideal driving rules for different surfaces.

Would the wheelchair recognize those surfaces and adjust accordingly?

There are two possibilities. Since you have a driver, the driver could change the setting when he or she encounters a new surface. That would be like standard shift in a car.

A second possibility we're pursuing is automatic shifting, which means the wheelchair has to have the ability to recognize when the terrain has changed. We had mounted on the wheelchair laser line stripe sensors that were developed by Carnegie Mellon University. It's a high-resolution sensor that can detect the important features of the terrain that can be used to identify changes in the terrain. It basically does terrain classifications. You could take the output of the terrain classifier that uses this sensor and use it to automatically adjust the control settings of the wheelchair.

When do you expect this could be available to wheelchair users?

My hope for its availability would be maybe five years. The actual wheelchair project as we're doing it now has only been going on for less than a year, but the technology development has been going on longer than that. Currently, we're starting to enter discussions with wheelchair companies on how to go about commercializing. The wheelchair development is not complete. We're still in the research stage. But we think we know how to do each component.

How did you come to do this work? Why is it important to you?

I was invited to participate in a center that's hosted at Carnegie Mellon University and the University of Pittsburgh. It's called the Quality of Life Technology Center. It's supported by the National Science Foundation. Years ago, I heard Rory Cooper at the University of Pittsburgh, a co-director of the center, give a talk on wheelchairs. After he gave his talk, I approached him because I saw a connection between what he wanted to do with wheelchairs and what we were doing with autonomous ground vehicles. That was my catalyst to do this.

My preference is to do more commercial-type research. I liked doing the research I do for the military, but I'd like to see other uses of it that can help the everyday person outside the military. There is a strong military connection to this research because it can benefit disabled veterans. I like that aspect and I also like the fact that the everyday consumer has the ability to use this technology.

Photo: Emmanuel Collins

This post was originally published on Smartplanet.com