Authors: Manish Prajapat, Johannes Köhler, Matteo Turchetta, Andreas Krause, Melanie N. Zeilinger

Published on: February 09, 2024

Impact Score: 8.07

Arxiv code: Arxiv:2402.06562

Summary

• What is new: First to guarantee exploration for non-linear systems with finite time sample complexity bounds, while ensuring safety with high probability.
• Why this is important: Exploring unknown environments safely restricts the autonomy of robots.
• What the research proposes: A novel framework, SageMPC, for safe, efficient exploration using Model Predictive Control with techniques like exploiting a Lipschitz bound, goal-directed exploration, and receding horizon style re-planning.
• Results: Demonstrated safe efficient exploration in challenging environments with a car model.

Technical Details

Technological frameworks used: SageMPC

Models used: Model Predictive Control

Data used: Non-linear systems dynamics, unknown environment constraints

Potential Impact

Automotive, delivery robots, autonomous drone scouting, robotic exploration and mapping companies

Want to implement this idea in a business?

We have generated a startup concept here: SageRobotics.