Safe Navigation using Neural Radiance Fields via Reachable Sets
Omanshu Thapliyal, Malarvizhi Sankaranarayanasamy, Ravigopal Vennelakanti
TLDR
This paper proposes a safe navigation method for robots in cluttered environments using reachable sets and NeRFs for obstacle representation.
Key contributions
- Integrates reachable sets to define safe navigation requirements for robots.
- Utilizes Neural Radiance Fields (NeRFs) for volumetric obstacle representation.
- Formulates path planning as a constrained optimal control problem with LMI constraints.
- Demonstrates safe navigation in simulations with numerous obstacles.
Why it matters
Autonomous systems need robust safe navigation. This work combines modern volumetric scene representation (NeRFs) with formal safety guarantees (reachable sets) and optimal control. It offers a promising approach for real-time safe path planning in complex, dynamic environments.
Original Abstract
Safe navigation in cluttered environments is an important challenge for autonomous systems. Robots navigating through obstacle ridden scenarios need to be able to navigate safely in the presence of obstacles, goals, and ego objects of varying geometries. In this work, reachable set representations of the robot's real-time capabilities in the state space can be utilized to capture safe navigation requirements. While neural radiance fields (NeRFs) are utilized to compute, store, and manipulate the volumetric representations of the obstacles, or ego vehicle, as needed. Constrained optimal control is employed to represent the resulting path planning problem, involving linear matrix inequality constraints. We present simulation results for path planning in the presence of numerous obstacles in two different scenarios. Safe navigation is demonstrated through using reachable sets in the corresponding constrained optimal control problems.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.