Development of an Automated Wheelchair Tiedown and Occupant Restraint System

Abstract

This project developed an automated wheelchair tiedown and restraint system (AWTORS) that could be safely and independently used in an automated vehicle (AV) by someone using a wheelchair as vehicle seating. The project used past research, computational modeling, prototype construction, volunteer evaluation, and dynamic testing to demonstrate feasibility. Computational modeling was used to optimize placement of the wheelchair station, locate the wheelchair anchorages relative to the occupant, optimize belt anchor locations, and determine airbag characteristics for front and side impacts. Frontal simulations showed improved injury measures with a Self-Conforming Rearseat AirBag (SCaRAB), particularly with suboptimal belt geometry. Side impact simulations showed adequate protection in nearside crashes with standard curtain airbags and outboard shoulder belt location. However, changes to belt geometry were insufficient to keep the occupant within the wheelchair during farside impacts, leading to design of a Center Airbag To Contain Humans (CATCH). Models of power and manual wheelchairs were developed and used to choose restraint and geometry parameters for sled testing. The concept for securing the wheelchair to the vehicle used hardware meeting specifications of the Universal Docking Interface Geometry (UDIG) that have been included in RESNA and ISO standards. Vehicle anchorages meeting the specifications were constructed, as were attachment designs for a commercial manual and power wheelchair. The occupant restraint portion of the AWTORS included an automatic seatbelt donning mechanism based on a past UMTRI prototype, but with geometric improvements. Volunteer testing was performed with 8 wheelchair users. Using the two study wheelchairs equipped with UDIG anchors, the study evaluated the usability of four wheelchair seating stations with different geometries, each with two different belt conditions. Data included videos of ingress and egress, scans of volunteer posture, and questionnaires to document the time spent docking the wheelchair and donning the seatbelt, belt fit, comfort, and potential usability issues. Average time for entry, docking, and donning was less than 2 minutes in all conditions. For three-quarters of trials, participants would recommend use of the docking and donning systems. The preparation of test fixtures for volunteer testing identified challenges in implementing optimal geometry defined through simulations. Ten frontal sled tests were performed to demonstrate differences with belt geometry and airbag presence, as well as to check the durability of UDIG anchors and attachments. Eight farside impacts were run to evaluate different versions of the CATCH bag, as well as to check durability of UDIG attachments in side impact.