Towards Interactive Crowd-aware Robot Navigation
混雑環境下における適応的なロボットナビゲーション
- Delivery
- Available on this site
- Format
- Price
- Non-members (tax incl.)：￥1,100 Members (tax incl.)：￥880
- Publication code
- 20224372
- Paper/Info type
- Journal of Society of Automotive Engineers of Japan
Vol.76 No.7
- Pages
- 122-128(Total 7 p)
- Date of publication
- Jul 2022
- Publisher
- JSAE
- Language
- Japanese
Detailed Information
| Category(J) | ホットトピックス Translation |
|---|---|
| Category(E) | Hot Topics |
| Author(J) | 1) 西村 真衣 |
| Author(E) | 1) Mai Nishimura |
| Affiliation(J) | 1) オムロンサイニックエックス |
| Abstract(J) | 混雑環境下における安全かつ効率的な経路計画はロボット工学の長年の課題である．本稿では，混雑環境におけるロボットナビゲーションのアプローチを俯瞰すると共に，深層強化学習により衝突回避と介入行動を効果的に切り替える適応的ナビゲーションの取り組みについて紹介する． Translation |
| Abstract(E) | Navigating robots in crowded places has been a long-standing challenge in robotics due to the severe requirements of high-level situational awareness for dynamic environments. In such“ human-centric” environments, robots are required to consider not only the efficiency of the planned path but also the safety to avoid potential collisions to nearby pedestrians. That is, interactive crowd-aware navigation involves the problem of safety-efficiency tradeoffs. Towards this goal, we introduce a deep reinforcement learning framework that balance the trade-offs inspired by well-known social dilemmas. |
