Abstract
Robots operating in human-centered environments must safely navigate among multiple dynamic obstacles to avoid collisions with people and surrounding infrastructure. Control barrier functions (CBFs) provide an effective mechanism for safety filtering, and recent CBF-based reinforcement learning (RL) methods embed such safety information into learned policies. However, executing only safety-filtered actions during training can restrict policy exploration, a limitation that becomes particularly consequential in dynamic scenes where safety depends on relative robot-obstacle motion. We propose DODGER, a safety-guided RL framework that directly executes policy-generated actions to drive training rollouts while using CBF-filtered references and constraint violations to shape the policy toward collision-avoidance behavior. We evaluate DODGER through a Dubins-car safety analysis and demonstrate goal-directed navigation among multiple dynamic obstacles in full-order humanoid simulation and real-world humanoid experiments using LiDAR-based perception, without a runtime safety filter.
Keywords
Subject
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Park, S., Lee, K., Kim, T., Lim, S., & Lee, Y. (2026). DODGER: Safety-Guided Reinforcement Learning for Robot Navigation Among Dynamic Obstacles. https://omanscience.com/en/articles/dodger-safety-guided-reinforcement-learning-for-robot-navigation-among-dynamic-obstacles
MLA 9
Park, Sanghyuk, et al. "DODGER: Safety-Guided Reinforcement Learning for Robot Navigation Among Dynamic Obstacles." https://omanscience.com/en/articles/dodger-safety-guided-reinforcement-learning-for-robot-navigation-among-dynamic-obstacles.
Chicago (author–date)
Park, Sanghyuk, Kwanwoo Lee, Taekyung Kim, Seohyeon Lim, and Yisoo Lee. 2026. "DODGER: Safety-Guided Reinforcement Learning for Robot Navigation Among Dynamic Obstacles." https://omanscience.com/en/articles/dodger-safety-guided-reinforcement-learning-for-robot-navigation-among-dynamic-obstacles.
Harvard
Park, S., Lee, K., Kim, T., Lim, S. and Lee, Y. (2026) 'DODGER: Safety-Guided Reinforcement Learning for Robot Navigation Among Dynamic Obstacles', Available at: https://omanscience.com/en/articles/dodger-safety-guided-reinforcement-learning-for-robot-navigation-among-dynamic-obstacles.
Vancouver
Park S, Lee K, Kim T, Lim S, Lee Y. DODGER: Safety-Guided Reinforcement Learning for Robot Navigation Among Dynamic Obstacles. https://omanscience.com/en/articles/dodger-safety-guided-reinforcement-learning-for-robot-navigation-among-dynamic-obstacles
IEEE
S. Park, K. Lee, T. Kim, S. Lim, and Y. Lee, "DODGER: Safety-Guided Reinforcement Learning for Robot Navigation Among Dynamic Obstacles," https://omanscience.com/en/articles/dodger-safety-guided-reinforcement-learning-for-robot-navigation-among-dynamic-obstacles.