Ethan Pedersen

Papers

Total Citations

H-Index

About

Ethan Pedersen is a rising researcher in autonomous robotics and intelligent agent design, with a focus on enabling robots to self-assess and adapt their behavior in real time. His work centers on two key areas: proficiency self-assessment (PSA) and multi-behavior decision-making. In his 2023 paper "Proficiency Self-Assessment without Breaking the Robot," Pedersen introduced assumption-alignment tracking (AAT), a novel method that allows robots to evaluate their own task performance by monitoring how well their internal assumptions match real-world conditions—without requiring risky or destructive testing. This approach has been cited 4 times and marks a significant step toward safer, more reliable autonomous systems. Complementing this, his paper "AlegAATr the Bandit" (2023, 4 citations) presents a bandit algorithm that dynamically selects among multiple pre-designed behaviors, enabling agents to adapt to changing tasks and environments. By integrating AAT with bandit-based behavior selection, Pedersen is pioneering frameworks that give robots both situational awareness and strategic flexibility. His work is particularly notable for its emphasis on safety and efficiency, addressing a critical gap in real-world robot deployment.