The Exoskeleton and Prosthetic Intelligent Controls (EPIC) Lab develops robotic prostheses and exoskeletons that use machine‑learning‑driven control systems to adapt assistance in real time based on biomechanical feedback and environmental context. Their end‑to‑end workflow spans custom hardware fabrication, bench‑top testing, controller optimization, and human performance evaluation to improve mobility, reduce user effort, and expand functional capabilities for individuals with limb loss or mobility impairments.
Funding
Funding not disclosed
Founders
Product
Problem
Individuals with limb loss or mobility impairments often rely on prosthetic limbs and exoskeletons that provide limited, pre‑programmed assistance, making it difficult to adapt to varying tasks, terrains, and user biomechanics. This lack of adaptability can reduce functional performance, increase user fatigue, and limit the devices’ usefulness in real‑world environments.
Solution
The EPIC Lab creates robotic prostheses and exoskeletons equipped with machine‑learning‑driven control systems that continuously adjust assistance levels based on real‑time biomechanical feedback and environmental context. By integrating sensor data, adaptive algorithms, and custom hardware, the devices can modulate torque, stiffness, and gait patterns to match the user’s intent and surroundings. The lab’s end‑to‑end workflow—from device fabrication and bench testing to controller optimization and human performance evaluation—ensures that each system is rigorously validated for safety and efficacy. This approach aims to enhance functional mobility, reduce effort, and expand the range of activities that users can perform with robotic assistive devices.
Target Audience
Primary users are individuals with lower‑limb amputation or neuromuscular disorders who require powered prosthetic or exoskeletal assistance, as well as clinicians and rehabilitation specialists evaluating advanced assistive technologies.
Features
- Custom‑fabricated prosthetic and exoskeleton hardware designed for modular sensor integration
- Real‑time biomechanical sensing (e.g., joint angles, ground reaction forces) to inform control decisions
- Machine‑learning algorithms that learn user‑specific movement patterns and adapt assistance dynamically
- Closed‑loop controller optimization pipeline linking bench‑top testing to human subject trials
- Comprehensive performance assessment tools measuring gait metrics, energy expenditure, and task success rates