Perceptual thresholds for virtual stiffness
With force-feedback gloves, users needed stiffness changes of 26–48% to reliably
distinguish virtual materials, well above the 8–22% reported for bare-finger
interaction. Weber fractions varied 1.8:1 across reference stiffness levels, violating
Weber’s Law, and discrimination accuracy depended on where in the workspace the hand
was. Measured with 23 participants using adaptive staircases and psychometric curve
fitting, these thresholds give designers concrete lower bounds for rendering perceivable
stiffness differences. Published in IEEE TVCG and presented at IEEE VR 2026.
How force feedback changes user behaviour
In a 52-participant experiment on virtual tool manipulation, force feedback reduced the
share of users who over-grip from 73% to 4% and cut applied grip force by 12–13%,
but did not improve task speed and increased cognitive load. Cluster analysis of the
behavioural data identified three distinct control strategies (optimal controllers 48%,
over-grippers 38%, variable controllers 13%), showing that one-size-fits-all haptic
rendering underserves most users and motivating user-adaptive interaction.
Iterating a VR rehabilitation system on user data
I designed and built a home VR hand-rehabilitation system that combines force-feedback
gloves with game mechanics, then iterated it across two user studies and assessment by
three physical therapists. Between system iterations, simulator sickness scores fell by
49% and user engagement improved, demonstrating the viability of home-based haptic
rehabilitation. Published at IEEE VR 2022.