pith. sign in

arxiv: 2212.05727 · v1 · pith:DAE66CUSnew · submitted 2022-12-12 · 💻 cs.LG · cs.RO

Evaluating Model-free Reinforcement Learning toward Safety-critical Tasks

classification 💻 cs.LG cs.RO
keywords safetyalgorithmslearningsafety-criticaltasksautonomousconstraintsevaluation
0
0 comments X
read the original abstract

Safety comes first in many real-world applications involving autonomous agents. Despite a large number of reinforcement learning (RL) methods focusing on safety-critical tasks, there is still a lack of high-quality evaluation of those algorithms that adheres to safety constraints at each decision step under complex and unknown dynamics. In this paper, we revisit prior work in this scope from the perspective of state-wise safe RL and categorize them as projection-based, recovery-based, and optimization-based approaches, respectively. Furthermore, we propose Unrolling Safety Layer (USL), a joint method that combines safety optimization and safety projection. This novel technique explicitly enforces hard constraints via the deep unrolling architecture and enjoys structural advantages in navigating the trade-off between reward improvement and constraint satisfaction. To facilitate further research in this area, we reproduce related algorithms in a unified pipeline and incorporate them into SafeRL-Kit, a toolkit that provides off-the-shelf interfaces and evaluation utilities for safety-critical tasks. We then perform a comparative study of the involved algorithms on six benchmarks ranging from robotic control to autonomous driving. The empirical results provide an insight into their applicability and robustness in learning zero-cost-return policies without task-dependent handcrafting. The project page is available at https://sites.google.com/view/saferlkit.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Scenario Generation for Risk-Aware Reinforcement Learning with Probably Approximately Safe Guarantees

    cs.LG 2026-06 unverdicted novelty 4.0

    Approximates encountered state distribution via VAE and constructs dual bound barrier certificates to provide probably approximately safe guarantees in RL by optimizing the non-robust region.