Associate professor of computer science and engineering John Spletzer is driven by a desire to see people with disabilities become more independent. And thanks to his drive, Spletzer and his team are leveraging technology developed for a driverless car to create a "smart" wheelchair that drives itself.
Spletzer competed in the 2007 DARPA Urban Challenge for robotic vehicles as part of a team from the University of Pennsylvania, Lehigh University and Lockheed Martin. Their Toyota Prius, fitted with laser and camera sensors, successfully navigated 58 miles of urban landscape on its own. DARPA, or the Defense Advanced Research Projects Agency, hopes to develop driverless ground-combat vehicles for the U.S. military to take soldiers out of harm's way.
While working with the robotic car, Spletzer saw an opportunity to translate how it navigates terrain and responds to its environment on the street, into how a robotic wheelchair navigates and responds to its environment on city sidewalks.
The sensors for the vehicle cost more than $150,000—too expensive for a cost-effective wheelchair. So Spletzer deployed the car, dubbed Little Ben, out onto the streets of Lehigh's South Bethlehem neighborhood. Imagine a process similar to the one Google uses to generate its online street view system. From the street, Little Ben uses LIDAR, or light detection and ranging lasers, to scan objects such as building facades, trees and parking meters. That data is used to generate 3D maps which the team downloads to the wheelchair, just as you might download a Google map to drive to a destination.
"We wanted a robot with a low cost and a high level of autonomy," said Spletzer, who runs the Vision, Assistive Devices, and Experimental Robotics (VADER) Laboratory. "Here the wheelchair benefits from Little Ben's expensive sensors, but without the high cost."
The robotic wheelchair has traversed a 1-km route and arrived at its destination to within an accuracy of 20 cm. During its trip, the robotic wheelchair learns from experience by comparing new objects it sees in the environment with images in its database.
Spletzer also took inspiration from another project, for which his lab provided technical and developmental support. This project, the Automated Transport and Retrieval System (ATRS), enables mobility independence by allowing wheelchair users to automatically stow and retrieve their chairs, and drive while sitting in traditional auto seats that meet federal safety regulations. The ATRS was commercialized by Freedom Sciences.