mlsanityData to 14 Oct 2026

Avoidance of an unexpected obstacle without reinforcement learning: Why not using advanced control-theoretic tools?

arXiv 2509.03721 ↗First posted 3 Sep 2025eess.SY cs.RO cs.SYcited by 0

On the map: Deep reinforcement learning methods → Safe and constrained RL

Abstract (arXiv)

This communication on collision avoidance with unexpected obstacles is motivated by some critical appraisals on reinforcement learning (RL) which "requires ridiculously large numbers of trials to learn any new task" (Yann LeCun). We use the classic Dubins' car in order to replace RL with flatness-based control, combined with the HEOL feedback setting, and the latest model-free predictive control approach. The two approaches lead to convincing computer experiments where the results with the model-based one are only slightly better. They exhibit a satisfactory robustness with respect to randomly generated mismatches/disturbances, which become excellent in the model-free case. Those properties would have been perhaps difficult to obtain with today's popular machine learning techniques in AI. Finally, we should emphasize that our two methods require a low computational burden.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Safe and constrained RL, one of 15 directions in Deep reinforcement learning methods. Builds on 14 of its 32 references that are in the map. First posted on arXiv 3 Sep 2025.

Builds on

14 of 32 references.

References · 14

  1. 2024Deep Reinforcement Learning for Robotics: A Survey of Real-World Successescited by 176
  2. 2018A Tour of Reinforcement Learning: The View from Continuous Controlcited by 151
  3. 2023Learning Agile Soccer Skills for a Bipedal Robot with Deep Reinforcement Learningcited by 133
  4. 2021Reinforcement Learning of the Prediction Horizon in Model Predictive Controlcited by 10
  5. 2020Manipulating Deformable Objects by Interleaving Prediction, Planning, and Controlcited by 9
  6. 2013Model-free controlcited by 7
  7. 2022Energy-based Legged Robots Terrain Traversability Modeling via Deep Inverse Reinforcement Learningcited by 5
  8. 2025Synchronization of Kuramoto oscillators via HEOL, and a discussion on AIcited by 3
  9. 2021An alternative to proportional-integral and proportional-integral-derivative regulators: Intelligent proportional-derivative regulatorscited by 2
  10. 2024Flatness-based control revisited: The HEOL settingcited by 2
  11. 2024The Euler-Lagrange equation and optimal control: Preliminary resultscited by 2
  12. 2014A model-free control strategy for an experimental greenhouse with an application to fault accommodationcited by 1
2 more
  1. 2025Model-Free Predictive Control: Introductory Algebraic Calculations, and a Comparison with HEOL and ANNscited by 1
  2. 2025Guidance and Control of Unmanned Surface Vehicles via HEOLcited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2024Dynamic Obstacle Avoidance through Uncertainty-Based Adaptive Planning with Diffusion
  2. 2024A Reinforcement Learning-Boosted Motion Planning Framework: Comprehensive Generalization Performance in Autonomous Driving
  3. 2026Composable Model-Free RL for Navigation with Input-Affine Systems
  4. 2022Automatically Learning Fallback Strategies with Model-Free Reinforcement Learning in Safety-Critical Driving Scenarios
  5. 2024Safe Reinforcement Learning of Robot Trajectories in the Presence of Moving Obstacles
  6. 2025Safe Reinforcement Learning with a Predictive Safety Filter for Motion Planning and Control: A Drifting Vehicle Example
  7. 2017Uncertainty-Aware Reinforcement Learning for Collision Avoidance
  8. 2021Offline Reinforcement Learning as Anti-Exploration