Reproduce “Baidu Apollo EM Motion Planner” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Haoyang Fan, Fan Zhu, Changchun Liu, Liangliang Zhang, Li Zhuang, Dong Li, Weicheng Zhu, Jiangtao Hu and 2 more
Baidu's Apollo EM Motion Planner generates safe and comfortable vehicle trajectories using an iterative Expectation-Maximization (EM)-type algorithm to refine path and speed profiles. This 'light-decision-based' approach processes a complete planning cycle in less than 100ms, demonstrated through extensive testing on Baidu Apollo autonomous vehicles.
Key insights
- The iterative Expectation-Maximization (EM)-type approach effectively addresses the circular dependency between path and speed planning, leading to converged, optimized solutions.
- A 'light-decision-based' framework allows basic decisions like follow, yield, or overtake to emerge naturally from the optimization process, enhancing adaptability without extensive predefined rules.
- Parallel evaluation of multiple candidate lanes enables more informed decisions for routing-driven or obstacle-avoidance lane changes.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.