Issue No. 2 - March, 1990

A Case Study: The Ultrasonic Head Controlled Wheelchair and Interface

David L. Jaffe, MS

The design, development, and technology transfer activities associated with the Ultrasonic Head Controlled Wheelchair (UHCW) has extended over the past ten years. It has involved my efforts as principal investigator as well as other researchers, students, test pilots, potential manufacturers, funding agencies, and personnel from the VA Rehabilitation R&D Evaluation Unit (REU) in Baltimore. This account briefly documents the history of bringing this design to commercial reality.

The student project
The UHCW project began in October, 1979 at the Design Division at Stanford University where a group of five graduate mechanical engineering students were assigned the task of designing an alternative and innovative control and guidance system for an electric vehicle capable of transporting quadriplegic individuals. Their nine month effort supervised by me and funded by RR&D Center resulted in "Smart Alec", a modified Everest and Jennings electric wheelchair outfitted with shaft encoders, electronics, a microcomputer, and numerous Polaroid ultrasonic distance ranging sensors. Two of these sensors monitored the user's head position for navigation while others detected obstacles in front of, behind, and to the sides of the wheelchair.

After the students graduated and went their separate ways, the project was moved to the RR&D Center where I continued to refine and publicize the design. In November of 1980, the project was awarded 4th prize in the Johns Hopkins Personal Computer Applications Competition in Washington, DC. At the System Sciences Conference in Honolulu, it was judged the best paper. And in March of 1981, the wheelchair was demonstrated at the ASME annual conference in Chicago.

Despite many demonstrations of this new mobility control concept, the students' solution was far from a practical vehicle. Its deficiencies were many. The method of loading the computer program from disk storage was both time consuming and could not be performed by a disabled user. The disk drives and CRT required a source of AC power during the loading phase, which might not be available in places where the chair might travel. The software proved difficult to maintain and modify. Although these problems could not be easily corrected on the existing prototype chair, the project did show the feasibility of computer controlled mobility directed by head position. It also showed that the ultrasonic sensor system could detect obstacles and take appropriate action. The practical execution of these concepts would have to be accomplished in a later design.

Second generation design
Since the ultrasonic head control technique appeared to have sufficient merit for further development, I decided to pursue a generalized wheelchair interface. A proposal was submitted to and funded by the Technology and Research Foundation of the Paralyzed Veterans of America for this work. In May, 1982 the Ultrasonic Head Control Unit project was begun at the RR&D Center, its goal being the development of a second generation wheelchair control system.

With that funding, advances in computer hardware, and new developments in high level programming languages, a new Ultrasonic Head Controlled Wheelchair was developed by the first half 1983. The computer's instructions were then permanently stored on a memory chip and the entire computer was powered by the wheelchair's batteries. A more accurate and faster technique of acquiring head position information improved the steering and control characteristics. In May, 1983 the first disabled user test drove the new design and in June it was demonstrated at the RESNA conference in San Diego. A dozen quadriplegic patients at the Palo Alto VA Spinal Cord Injury Service subsequently tested the prototype.

In response to many technical inquires about the wheelchair, I compiled a sixty-five page Technical Manual. It contains complete descriptions of the workings of the UHCW, including schematics, wiring diagrams, computer program source code, and parts list. With this document a knowledgeable engineer in an adequately equipped laboratory could duplicate the head controlled wheelchair. In the past five years over one hundred investigators, researchers, and interested manufacturers world-wide have received copies of the manual.

In 1985, as work was progressing on a voice controlled mobile robot, modifications to the wheelchair-mounted head control unit were made to enable its use in specifying movement trajectories. This capability would allow the severely disabled user to control complex robot movements using only head motions.

Despite continuing interest in the wheelchair and ultrasonic distance ranging technology, no manufacturer had stepped forward to pursue commercial development of the design. In mid-1985, the Rehabilitation R&D Evaluation Unit (REU) was formed within the VA system. (For a description of REU, see the related article in this newsletter.) A Request for Evaluation for the UHCW from RR&D was one of the first to be received and approved by this new agency.

The evaluation contract
The proposal called for the development and delivery of four pre-production UHCW devices. These units were to be clinically tested at VA Medical Centers around the country for certification. It was not until 1987 that the funds were received and contracted to Eureka Laboratories of Sacramento, CA as the result of a competitive bidding process. The company was initially made aware of the RR&D Center's work during a Manufacturer's Conference organized jointly by the VA and the Department of Commerce. In one-year's time Eureka completely redesigned the UHCW, taking advantage of new microcontroller technology and incorporating safety and design features necessary of a production device. A packaging upgrade to improve the aesthetics was also accomplished. Their initial device underwent a successful local acceptance test in June, 1989 and was delivered to REU shortly thereafter.

Under the REU program, clinical evaluation is to be accomplished at several VA facilities. A protocol has been developed by REU encompassing testing methods, questionnaires, subject selection, and data collection methods. The process is estimated to take about one year and if successful, will result in an approval for VA physicians to prescribe the UHCW for quadriplegic veterans. By virtue of their early involvement in this project, Eureka will be in a good position to fill any orders for UHCW's submitted by the VA or other third-party payers.

While the events described above may not be typical or even necessary for other projects, they do show what it takes bring a good project idea to the point of commercialization: a concerted, long-term technical effort, several infusions of funding, and plenty of patience on the part of RR&D Center investigators. This case also illustrates the advantage to all parties when manufacturers become aware of projects in their early stages. The R&D team, the manufacturers and the end user all benefit from this early involvement.