Image: The Heterogeneous Robotics Research Lab (HeRoLab) at the University of Georgia, led by Prof. Ramviyas Parasuraman, is pleased to announce that five papers have been accepted for oral presentation at the IEEE International Conference on Robotics and Automation (ICRA) 2026, the top-ranked and highly selective international conference in Robotics, and listed in CSRankings.org. ICRA 2026 will be held in Vienna, Austria, in June 2026. These acceptances highlight HeRoLab’s ongoing contributions to advancing robotics and multi-robot systems, with innovations spanning GPS-denied autonomy, multi-robot coordination, safety-aware human–robot interaction, explainable artificial intelligence, and energy-efficient persistent autonomy. Accepted Papers and Research Highlights 1) DCL-Sparse: Distributed Relative Localization in Sparse Graphs Authors: Atharva Sagale (SoC PhD Student), Tohid Kargar Tasooji (Postdoc in SoC), Ramviyas Parasuraman (Associate Professor in SoC) This paper presents a new distributed localization approach that enables robot teams to accurately determine their positions in GPS-denied environments with sparse and noisy connectivity. The proposed method significantly improves robustness and scalability, achieving up to a 95% reduction in localization error compared to state-of-the-art techniques. 2) Multi-Robot Informative Sampling and Coverage in GPS-Denied Environments (Conference forthcoming) Authors: Aiman Munir (SoC PhD Alumni), Ehsan Latif (SoC PhD Alumni), Ramviyas Parasuraman (Associate Professor in SoC) This work introduces a framework that allows robot teams to explore, map, and collect informative data without relying on GPS. By balancing informative sampling and spatial coverage under localization uncertainty, the approach outperforms alternative GPS-denied methods by up to 54%, while achieving performance comparable to GPS-based systems. 3) Imitation-BT: Automating Behavior Tree Generation by Echoing Reinforcement Learning Agents Authors: Shailendra Sekhar Bathula (MSAI Alumni), Ramviyas Parasuraman (Associate Professor in SoC) This paper addresses the challenge of deploying opaque reinforcement learning policies in safety-critical systems. The proposed method automatically converts learned policies into transparent, interpretable behavior trees, retaining high task performance while significantly improving explainability, trust, and verification ease. 4) Energy-Aware Informative Path Planning for Heterogeneous Multi-Robot Systems (Conference forthcoming) Authors: Aiman Munir (SoC PhD Alumni), Ayan Dutta (Associate Professor, University of North Florida), Ramviyas Parasuraman (Associate Professor in SoC) This work presents an energy-aware planning framework for long-term, persistent multi-robot deployments. By explicitly accounting for robot energy states and task uncertainty, the method enables efficient transitions between exploration, exploitation, and recharging, achieving up to 32% energy savings while maintaining high-quality environmental reconstructions. 5) FRESHR-GSI: A Generalized Safety Model and Evaluation Framework for Mobile Robots in Multi-Human Environments Authors: Pranav Pandey (SoC PhD Student), Ramviyas Parasuraman (Associate Professor in SoC), Prashant Doshi (Professor in SoC) This paper introduces a new framework for quantifying human safety around mobile robots operating in shared spaces. By incorporating human–robot distance, velocity, orientation, and other dynamic factors, the approach delivers more accurate and responsive safety assessments than existing models, supporting safer deployment of robots in public and human-centered environments.Additional details for the papers, including preprints, videos, and open-source code, are available at:https://hero.uga.edu/publications Acknowledgments We gratefully acknowledge the resources and support from the School of Computing and extend our congratulations to the students and alumni of the HeRoLab, as well as external collaborators, for their contributions to this work. This research was supported by the funding agencies, including NIFA, ARL, and DEVCOM, whose support is essential to the success of this research.The team looks forward to presenting these results at ICRA 2026 in Vienna and engaging with the international robotics research community. Type of News/Audience: Research
