Sensorless Foot-Tip Deceleration and Ground-Contact Reachability for a Fall-Prevention Support Mechanism in Smart Walkers

A lateral support mechanism that rapidly lowers a support foot can extend the base of support of a smart walker during a loss of balance, but rapid deployment raises the velocity at which the foot reaches the ground; this touchdown velocity must be reduced, and the contact position anticipated, without adding a force or contact sensor. This study develops a sensorless, uncertainty-aware predictive deceleration strategy that uses the kinematic model and actuator-position information alone, with no dedicated contact sensor and no dynamic observer. Monte Carlo propagation of joint clearance quantifies the contact-position uncertainty, which is combined with an assumed actuation delay to determine the predictive-deceleration onset. Across joint clearances of 0.10–0.50 mm, the mean contact position shows no systematic shift, while the uncertainty half-width of the central 99% interval increases from 0.22 to 1.09 mm. Over the investigated target-touchdown-velocity and actuation-delay ranges, all cases satisfy the most restrictive 350 ms budget, with a maximum total deployment time of approximately 314 ms, and therefore, also satisfy the less restrictive timing references; in all sampled geometries, ground contact remains reachable, with a smallest margin of 0.82 mm at the largest investigated clearance. These numerical results show that clearance-induced contact-position uncertainty can be quantified and incorporated directly into predictive deceleration, and experimental validation is the planned next stage.