Research Overview

The Bionic Systems Lab at UCR investigates theoretical and engineering foundations of human-robot systems aimed at restoring and enhancing human mobility and perception, including wearable and assistive robots, active prostheses and orthoses, and neuroprostheses. Our approach includes three synergistic areas: (1) design principles for human-centric and body-mounted robots, (2) mathematical models and control algorithms for human-robot systems, and (3) human-subject studies and data-based modeling of sensorimotor behavior. We envision wearable and assistive robots seamlessly integrating with users to empower people with disabilities, improve precision and personalized medicine, and extend sensory experience and physicality.



Fabric-Based Soft Wearable Robots


Rigid wearable robots limit the natural workspace of limbs, are prone to kinematic incompatibilities, and add substantial inertia, thereby requiring compensatory nonphysiological muscle strategies during movement. In contrast, soft wearable robots are lightweight, compliant, portable, and easy to don/doff, enhancing the potential for daily use and at home rehabilitation. Using a variety of textiles, fabric-based pneumatic soft actuators can be created that are mechanically programmed to bend, twist, and contract when inflated. Antagonistic configurations allow independent stiffness and equilibrium control for active joint assistance. Arranging actuators in series and parallel could form functional robotic apparel for movement assistance and strength augmentation.

Nonlinear Elastic Actuators



Biology harnesses passive tissues, like tendons and ligaments, to effectively store and transfer energy. Likewise, incorporating passive energy storing structures can minimize weight and size of lower-limb prostheses (or orthoses). However, to achieve these benefits, the passive element(s) must be carefully tuned. By personalizing the passive loading response of an active ankle-foot prosthesis the motor requirements can be largely decreased. The innovation involves optimizing a nonlinear spring to match the user specific stiffness trajectory, then realizing the nonlinearity with a novel cam mechanism. The result, a so-called nonlinear elastic actuator, is a first step towards the broader goal of replicating locomotion using mostly passive structures.


Human-in-the-Loop Optimization


Human-in-the-loop optimization, where robot behavior is adjusted based on the combined human-robot output, offers a framework for tuning robot parameters. The critical challenges include ensuring stable behavior in the presence of time-varying human dynamics and properly exploiting the new information after observing the combined human-robot output, e.g., to converge towards an good control policy. A key area of our research explores iterative learning control to improve human-robot augmentation.

Sensory Supplementation



Interaction forces, detected by mechanoreceptors in the skin, provide valuable sensory information during locomotion. However, current prostheses have little direct feedback to the user. Our research seeks to design systems that produce haptic feedback for prostheses users: interaction forces sensed at the prosthetic device are relayed to the user through a set of vibrotactile actuators worn around the residual limb. The hope is that the feedback will be incorporated into the user's sensorimotor response. Additional applications include sensory substitution. For example, a white cane equipped with sonar can provide haptic feedback to the blind.