Texas Tech University

Advancing Robot Learning: Assistant Professor's Research Accepted at ICRA 2024

stas

Assistant Professor has had two research papers on robot learning accepted at ICRA 2024

Stas Tiomkin, assistant professor at the Department of Computer Science in the Edward E. Whitacre College of Engineering  at Texas Tech University has had two research papers on robot learning accepted at the International Conference on Robotics and Automation (ICRA).

ICRA ranked as the #1 conference in robotics by Google Scholar, showcases groundbreaking advancements in the field. Tiomkin’s work, along with his collaborators, pushes the boundaries of robot learning, acoustic wave manipulation, and reward design for autonomous systems.

Paper 1: "Acoustic Wave Manipulation Through Sparse Robotic Actuation"
Authors: Tristan Shah*, Noam Smilovich*, Samer Gerges, Feruza Amirkulova, Stas Tiomkin (corresponding author)
(*PhD students at TTU, equal contribution)
Key Idea: This work explores how robots can manipulate acoustic waves—a novel problem with applications in artificial materials, ultrasonic tools, and energy harvesting. The proposed data-driven method surpasses existing learning-based approaches in both solution quality and computational efficiency.

Paper 2: "SuPLE: Robot Learning with Lyapunov Rewards"
Authors: Phu Nguyen, Daniel Polani, Stas Tiomkin (corresponding author)
 Key Idea: This research introduces SuPLE (Sum of Positive Lyapunov Exponents) as a system-inherent reward function for robot learning, eliminating the need for external assumptions. This approach paves the way for more efficient and generalizable robot learning strategies.

Congratulations to Dr. Tiomkin and his collaborators on this remarkable achievement!