Papers
28
Total Citations
926
H-Index
18
About
Jennie Si is a pioneering researcher at the intersection of reinforcement learning, adaptive dynamic programming, and wearable robotics, with particular expertise in developing intelligent control systems for robotic prostheses and exoskeletons. Her most impactful work centers on automating the personalization of lower limb assistive devices—a notoriously difficult problem given the vast number of control parameters that must be tuned for individual users. Her landmark 2019 study on online reinforcement learning for robotic knee prosthesis personalization (186 citations) demonstrated that human-in-the-loop RL could circumvent the limitations of traditional control design, transforming how clinicians and engineers approach device fitting. Building on this, Si has advanced actor-critic architectures, flexible policy iteration, and adaptive dynamic programming frameworks to make personalization faster, more data-efficient, and theoretically grounded in stability and optimality guarantees. Her work spans powered knee prostheses, hip exoskeletons, and gait symmetry analysis, offering insights into wearer-prosthesis interaction that were previously inaccessible. With a body of work accumulating over 600 citations across these papers alone, Si's contributions are reshaping rehabilitation robotics, making intelligent wearable devices meaningfully responsive to the humans who depend on them.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4Supervised Actor-Critic Reinforcement Learning69 citations · 2012
- 5
- 6
- 7
- 8
- 9
- 10