AS-Gaussian dynamics improve precision force training in VR haptics

Researchers led by Alberto Garzás-Villar (arXiv preprint 2605.26782) investigated how manipulating the rendered dynamics of tangible virtual objects can improve training of fine force generation, a key component in post-stroke neurorehabilitation. Using a robotic haptic device combined with VR, 50 healthy participants performed a curling-inspired task: stretching a virtual spring to generate a target release force to propel a stone to a predefined location. The spring's force-elongation relationship was modeled as linear, Gaussian, or antisymmetric Gaussian (AS-Gaussian), with the latter having zero derivative at the target force.

The study found that the AS-Gaussian group consistently achieved higher force accuracy during training compared to the linear group. The Gaussian group only outperformed the linear group toward the end of training. Personality traits influenced outcomes: higher Free Spirit scores correlated with poorer performance and less exploration under Gaussian dynamics, while higher Transform-of-Challenge scores correlated with more exploration. Despite these training effects, no significant differences in long-term retention emerged across spring types or personality traits. Notably, participants primarily relied on learned target elongation rather than target force, as shown in a transfer task with different stiffness but same target force. The authors conclude the method is promising for somatosensory neurorehabilitation but requires refinement to reduce reliance on proprioceptive cues before testing with neurological patients.