Max Robotics — Global Service Robot Directory + US Certification

About

Shreyas Sundara Raman is a rising researcher at the intersection of large language models (LLMs) and robotics, whose work focuses on making LLM-driven robot agents both smarter and safer. His major contributions center on two critical challenges: enabling robots to recover from failures intelligently, and ensuring they operate within safe constraints. In his highly cited work "CAPE: Corrective Actions from Precondition Errors using Large Language Models" (accumulating 22 citations across its versions), Raman pioneered a method that moves beyond simply retrying failed actions. Instead, his approach uses LLMs to diagnose the underlying cause of a failure—such as a missing precondition—and then generates a corrective action to resolve it, giving robots true error-recovery capabilities. Complementing this, his paper "Plug in the Safety Chip: Enforcing Constraints for LLM-driven Robot Agents" (34 citations) addresses the critical need for safety in autonomous systems, introducing a framework to enforce hard constraints on LLM-generated plans. This work has been recognized as a vital step toward deploying LLM agents in real-world, safety-critical environments. Raman’s research is shaping the future of reliable and responsible autonomous robotics.