Identifying Comfort Areas in 3D Space for Persons with Upper Extremity Mobility Impairments Using Virtual Reality
Shanmugam Muruga Palaniappan, Ting Zhang, Bradley S. Duerstock · 2019 · Proceedings of the 21st International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS) · doi:10.1145/3308561.3353810
Summary
This short paper presents a method for using virtual reality exergaming to identify personalized comfort areas in 3D space for people with upper extremity mobility impairments (UEMIs), enabling optimal placement of assistive technology input devices like switches, joysticks, and touchpads. The current approach to positioning assistive devices relies on occupational therapists conducting iterative trial-and-error experiments with physical prototypes — a process that is expensive, time-consuming, and exhausting for clients with conditions like spinal cord injury, stroke, multiple sclerosis, Parkinson's disease, or ALS. The proposed method uses a VR exergame where users wear an HTC Vive headset and interact with virtual spheres arranged in rings around them, pushing spheres outward as far as possible within a 2-minute period. The hand motion data collected during this game is then analyzed using Gaussian kernel density estimation to identify areas of frequent motion (comfort zones) and infrequent motion (discomfort zones). The highest-density cluster is identified as the resting position (usually on the lap) and excluded from comfort analysis. K-means clustering (k=2) is applied to the top 30% density data (comfort areas) and bottom 2% density data (least comfort areas), producing four distinct zones. The comfort levels are validated biomechanically by computing the static shoulder joint torques required to reach each zone using a Denavit-Hartenberg robotic arm model in MATLAB.
Key findings
A case study with one male participant with C4/C5 tetraplegia validated the approach. The shoulder torque calculations confirmed that comfort areas required significantly less force — on average 47.5% lower torque than least comfort areas. The participant then completed a typing task using a two-button step-scan keyboard (Blue2 Bluetooth switch from AbleNet on an iPad) with the switch positioned at each of the four identified zones. At comfort areas, the participant typed eight-letter words in approximately 40 seconds with 100% accuracy. At least comfort areas, typing time roughly doubled (~80-100 seconds) and accuracy dropped to approximately 50%, with half of trials containing at least one error. Subjective survey ratings aligned with the objective data: the comfort-right zone (closest to the lap resting position) received the highest comfort and ease-of-reach ratings (5/5) and lowest frustration (1/5), while the least-comfort-right zone received inverse scores. The mixed reality visualization — rendering the identified comfort and discomfort zones as colored virtual spheres overlaid on the real user — enabled rapid verification of the extracted zones relative to the user's body.
Relevance
This research demonstrates how virtual reality can transform the traditionally slow, expensive, and physically demanding process of setting up assistive technology for people with severe motor impairments. For individuals with tetraplegia or other UEMIs, the placement of input devices (switches, joysticks, sip-and-puff controllers) directly determines their ability to use computers, control wheelchairs, operate environmental controls, and interact with the world. Even small differences in device placement can dramatically affect performance — as shown by the doubling of typing time and halving of accuracy when the switch was moved from a comfort zone to a discomfort zone. Currently, this positioning is done through clinical trial-and-error, which is fatiguing for clients and difficult to reproduce across settings. The VR approach collects comprehensive motion data in a brief, game-like session that is more engaging than clinical assessments. For accessibility practitioners and rehabilitation engineers, the method provides an objective, data-driven alternative to subjective placement decisions. The applications extend beyond switch positioning to wheelchair control customization, vehicle modification design, human-robot collaboration workspace planning, and any scenario where assistive interfaces must be placed within a person's individualized comfort zone rather than at standardized locations.
Tags: virtual reality · motor disability · spinal cord injury · ergonomics · assistive technology · workspace design · switch access · tetraplegia · upper extremity impairment · personalization