Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2509.17204 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909801652420608 |
|---|---|
| author | Han, James R. Vanniasinghe, Mithun Sahak, Hshmat Rhinehart, Nicholas Barfoot, Timothy D. |
| author_facet | Han, James R. Vanniasinghe, Mithun Sahak, Hshmat Rhinehart, Nicholas Barfoot, Timothy D. |
| contents | Scaling Reinforcement Learning to in-the-wild social robot navigation is both data-intensive and unsafe, since policies must learn through direct interaction and inevitably encounter collisions. Offline Imitation learning (IL) avoids these risks by collecting expert demonstrations safely, training entirely offline, and deploying policies zero-shot. However, we find that naively applying Behaviour Cloning (BC) to social navigation is insufficient; achieving strong performance requires careful architectural and training choices. We present Ratatouille, a pipeline and model architecture that, without changing the data, reduces collisions per meter by 6 times and improves success rate by 3 times compared to naive BC. We validate our approach in both simulation and the real world, where we collected over 11 hours of data on a dense university campus. We further demonstrate qualitative results in a public food court. Our findings highlight that thoughtful IL design, rather than additional data, can substantially improve safety and reliability in real-world social navigation. Video: https://youtu.be/tOdLTXsaYLQ. Code will be released after acceptance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_17204 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Ratatouille: Imitation Learning Ingredients for Real-world Social Robot Navigation Han, James R. Vanniasinghe, Mithun Sahak, Hshmat Rhinehart, Nicholas Barfoot, Timothy D. Robotics Scaling Reinforcement Learning to in-the-wild social robot navigation is both data-intensive and unsafe, since policies must learn through direct interaction and inevitably encounter collisions. Offline Imitation learning (IL) avoids these risks by collecting expert demonstrations safely, training entirely offline, and deploying policies zero-shot. However, we find that naively applying Behaviour Cloning (BC) to social navigation is insufficient; achieving strong performance requires careful architectural and training choices. We present Ratatouille, a pipeline and model architecture that, without changing the data, reduces collisions per meter by 6 times and improves success rate by 3 times compared to naive BC. We validate our approach in both simulation and the real world, where we collected over 11 hours of data on a dense university campus. We further demonstrate qualitative results in a public food court. Our findings highlight that thoughtful IL design, rather than additional data, can substantially improve safety and reliability in real-world social navigation. Video: https://youtu.be/tOdLTXsaYLQ. Code will be released after acceptance. |
| title | Ratatouille: Imitation Learning Ingredients for Real-world Social Robot Navigation |
| topic | Robotics |
| url | https://arxiv.org/abs/2509.17204 |