# HASARD: A Benchmark for Vision-Based Safe Reinforcement Learning in Embodied Agents

arXiv: https://arxiv.org/abs/2503.08241 · HTML: https://mlsanity.com/paper/2503.08241/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Placed in RL for robot locomotion and manipulation, one of 15 directions in Deep reinforcement learning methods.
- Builds on 45 of its 64 references that are in the map.
- First posted on arXiv 11 Mar 2025.

## Placed in

- Deep reinforcement learning methods / [RL for robot locomotion and manipulation](https://mlsanity.com/deeprl/rl-for-robot-locomotion-and-manipulation/)

## Builds on (most cited first, 30 of 45)

- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [Playing Atari with Deep Reinforcement Learning](https://mlsanity.com/paper/1312.5602/) (2013)
- [Trust Region Policy Optimization](https://mlsanity.com/paper/1502.05477/) (2015)
- [CARLA: An Open Urban Driving Simulator](https://mlsanity.com/paper/1711.03938/) (2017)
- [Dota 2 with Large Scale Deep Reinforcement Learning](https://mlsanity.com/paper/1912.06680/) (2019)
- [robosuite: A Modular Simulation Framework and Benchmark for Robot Learning](https://mlsanity.com/paper/2009.12293/) (2020)
- [Safe Reinforcement Learning via Shielding](https://mlsanity.com/paper/1708.08611/) (2017)
- [Minigrid & Miniworld: Modular & Customizable Reinforcement Learning Environments for Goal-Oriented Tasks](https://mlsanity.com/paper/2306.13831/) (2023)
- [Responsive Safety in Reinforcement Learning by PID Lagrangian Methods](https://mlsanity.com/paper/2007.03964/) (2020)
- [Diffusion Models Are Real-Time Game Engines](https://mlsanity.com/paper/2408.14837/) (2024)
- [MetaDrive: Composing Diverse Driving Scenarios for Generalizable Reinforcement Learning](https://mlsanity.com/paper/2109.12674/) (2021)
- [StarCraft II: A New Challenge for Reinforcement Learning](https://mlsanity.com/paper/1708.04782/) (2017)
- [Reverse Curriculum Generation for Reinforcement Learning](https://mlsanity.com/paper/1707.05300/) (2017)
- [A Review of Safe Reinforcement Learning: Methods, Theory and Applications](https://mlsanity.com/paper/2205.10330/) (2022)
- [ViZDoom: A Doom-based AI Research Platform for Visual Reinforcement Learning](https://mlsanity.com/paper/1605.02097/) (2016)
- [Safety-Gymnasium: A Unified Safe Reinforcement Learning Benchmark](https://mlsanity.com/paper/2310.12567/) (2023)
- [AI Safety Gridworlds](https://mlsanity.com/paper/1711.09883/) (2017)
- [Stabilizing Deep Q-Learning with ConvNets and Vision Transformers under Data Augmentation](https://mlsanity.com/paper/2107.00644/) (2021)
- [Sample Factory: Egocentric 3D Control from Pixels at 100000 FPS with Asynchronous Reinforcement Learning](https://mlsanity.com/paper/2006.11751/) (2020)
- [Driving Policy Transfer via Modularity and Abstraction](https://mlsanity.com/paper/1804.09364/) (2018)
- [Leave no Trace: Learning to Reset for Safe and Autonomous Reinforcement Learning](https://mlsanity.com/paper/1711.06782/) (2017)
- [MindAgent: Emergent Gaming Interaction](https://mlsanity.com/paper/2309.09971/) (2023)
- [OmniSafe: An Infrastructure for Accelerating Safe Reinforcement Learning Research](https://mlsanity.com/paper/2305.09304/) (2023)
- [An empirical investigation of the challenges of real-world reinforcement learning](https://mlsanity.com/paper/2003.11881/) (2020)
- [NeuralField-LDM: Scene Generation with Hierarchical Latent Diffusion Models](https://mlsanity.com/paper/2304.09787/) (2023)
- [Safe Reinforcement Learning via Curriculum Induction](https://mlsanity.com/paper/2006.12136/) (2020)
- [Saute RL: Almost Surely Safe Reinforcement Learning Using State Augmentation](https://mlsanity.com/paper/2202.06558/) (2022)
- [Penalized Proximal Policy Optimization for Safe Reinforcement Learning](https://mlsanity.com/paper/2205.11814/) (2022)
- [Deep Reinforcement Learning for Traffic Light Control in Vehicular Networks](https://mlsanity.com/paper/1803.11115/) (2018)
- [safe-control-gym: a Unified Benchmark Suite for Safe Learning-based Control and Reinforcement Learning in Robotics](https://mlsanity.com/paper/2109.06325/) (2021)

## Built on by (most cited first, 6 of 6)

- [Safety in Embodied AI: A Survey of Risks, Attacks, and Defenses](https://mlsanity.com/paper/2605.02900/) (2026)
- [CRAX: Fast Safe Reinforcement Learning Benchmarking](https://mlsanity.com/paper/2606.20376/) (2026)
- [Perturbation-mitigated USV Navigation with Distributionally Robust Reinforcement Learning](https://mlsanity.com/paper/2512.00030/) (2025)
- [Isolation as a First-Class Principle for LLM-Agent System Safety: Concepts, Taxonomy, Challenges and Future Directions](https://mlsanity.com/paper/2607.12406/) (2026)
- [Evaluation Metrics for Safe Reinforcement Learning](https://mlsanity.com/paper/2609.15315/) (2026)
- [Factoriax: A GPU-Accelerated Factorio-Style Simulator for Reinforcement Learning](https://mlsanity.com/paper/2610.05569/) (2026)

## Abstract (arXiv)

Advancing safe autonomous systems through reinforcement learning (RL) requires robust benchmarks to evaluate performance, analyze methods, and assess agent competencies. Humans primarily rely on embodied visual perception to safely navigate and interact with their surroundings, making it a valuable capability for RL agents. However, existing vision-based 3D benchmarks only consider simple navigation tasks. To address this shortcoming, we introduce HASARD, a suite of diverse and complex tasks to HArness SAfe RL with Doom, requiring strategic decision-making, comprehending spatial relationships, and predicting the short-term future. HASARD features three difficulty levels and two action spaces. An empirical evaluation of popular baseline methods demonstrates the benchmark's complexity, unique challenges, and reward-cost trade-offs. Visualizing agent navigation during training with top-down heatmaps provides insight into a method's learning process. Incrementally training across difficulty levels offers an implicit learning curriculum. HASARD is the first safe RL benchmark to exclusively target egocentric vision-based learning, offering a cost-effective and insightful way to explore the potential and boundaries of current and future safe RL methods. The environments and baseline implementations are open-sourced at https://sites.google.com/view/hasard-bench/.