SMARTS Simulation Platform
Jump to navigation
Jump to search
A SMARTS Simulation Platform is a multi-agent simulation platform that focuses on research in autonomous driving.
- AKA: Scalable Multi-Agent Reinforcement Learning Training School.
- Context:
- It can (typically) be a part of the XingTian suite of RL platforms from Huawei Noah's Ark Lab.
- It can (typically) be is designed to generate realistic and diverse Agent Interactions.
- It supports the training, accumulation, and use of diverse behavior models of road users.
- It can provide realistic and diverse interactions to enable deeper and broader research on multi-agent interaction.
- It can be designed to address the challenge of how to interact competently with diverse road users in various scenarios, a key unsolved problem in the field of autonomous driving.
- ...
- Example(s):
- ...
- Counter-Example(s):
- See: Driving SMARTS 2.0 Competition, Autonomous Driving, Reinforcement Learning, Simulation Platform, Multi-Agent System, [[
References
2020
- (Zhou et al., 2020) ⇒ Zhou, Ming, Jun Luo, Julian Villella, Yaodong Yang, David Rusu, Jiayu Miao, Weinan Zhang et al. “Smarts: Scalable multi-agent reinforcement learning training school for autonomous driving." arXiv preprint arXiv:2010.09776 (2020).
- ABSTRACT: Multi-agent interaction is a fundamental aspect of autonomous driving in the real world. Despite more than a decade of research and development, the problem of how to competently interact with diverse road users in diverse scenarios remains largely unsolved. Learning methods have much to offer towards solving this problem. But they require a realistic multi-agent simulator that generates diverse and competent driving interactions. To meet this need, we develop a dedicated simulation platform called SMARTS (Scalable Multi-Agent RL Training School). SMARTS supports the training, accumulation, and use of diverse behavior models of road users. These are in turn used to create increasingly more realistic and diverse interactions that enable deeper and broader research on multi-agent interaction. In this paper, we describe the design goals of SMARTS, explain its basic architecture and its key features, and illustrate its use through concrete multi-agent experiments on interactive scenarios. We open-source the SMARTS platform and the associated benchmark tasks and evaluation metrics to encourage and empower research on multi-agent learning for autonomous driving.