AGIBOT

机器人具身算法工程师(VLA/RL)

AGIBOT  •  Onsite  •  2 hours ago
Apply
AI can make mistakes so check important info. Chat history is never stored.

Job Description

机器人具身算法工程师(VLA/RL)

上海

全职

职位描述

1、VLA 端到端模型研发: 设计并训练端到端视觉-语言-动作模型(如 RT-2、OpenVLA、π0、Diffusion Policy、3D Diffusion Actor 等),实现从视觉输入与语言指令到机器人低维动作空间的直接映射; 2、研究动作表示与生成机制: 动作 Token 化(Action Tokenization)、隐动作量化(Latent Action Quantization, LAPA)、扩散式动作生成(Diffusion Policy)、流匹配(Flow Matching)等前沿方案; 3、探索 RL 与模仿学习(IL)的混合训练范式:利用人类演示数据初始化策略,再通过 RL 进行微调和鲁棒性增强; 4、强化学习运控与策略优化: 包括 PPO、SAC、RLHF(Human Feedback for Robotics)、DAPG 等算法在关节级/任务级控制中的应用; 5、模型部署与实时性优化: 负责 VLA/RL 模型在机器人芯片(Orin、Thor)上的轻量化部署,满足实时控制需求(端到端延迟 < 50ms,控制频率 ≥ 50Hz),解决长上下文视觉序列与语言指令的并行处理瓶颈; 6、设计模型与底层运控(WBC/MPC)的协同接口: 实现高层 VLA 策略输出与低层全身控制器的平滑衔接,确保物理安全性与稳定性。

职位要求

1、计算机科学、人工智能、机器人学、自动化、控制理论等相关专业,硕士及以上学历; 2、VLA 深度经验:精通至少一种端到端 VLA 架构(RT-2、OpenVLA、Octo、π0、GR00T 等),理解其动作空间设计、训练目标与推理机制; 3、强化学习:熟练掌握 PPO、SAC、RLHF、GAE、TD(λ) 等算法原理与实现,具备在连续动作空间(机器人控制)中的 RL 训练经验; 4、机器人学基础:熟悉运动学(FK/IK)、动力学、坐标系变换(SMPL Y-up/Z-up 等)、关节空间与任务空间映射、全身控制(WBC)概念; 5、工程能力:精通 PyTorch,熟悉 CUDA 优化、模型服务化(vLLM、Triton)优先; 6、仿真与部署:熟练使用 Isaac Sim、MuJoCo、Genesis 等仿真环境;具备真实机器人平台(人形/四足/机械臂)的部署与调试经验。

投递

AGIBOT

About AGIBOT

AGIBOT is an Embodied AI foundation model company developing both the intelligence layer and the corresponding robotic embodiments needed to bring general intelligence into the physical world. Founded in 2023, AGIBOT is building a full-stack Embodied AI platform across foundation models, data infrastructure, simulation, robotic hardware, and deployment. AGIBOT’s “Three Intelligences in One” architecture integrates Locomotion Intelligence, Interaction Intelligence, and Manipulation Intelligence into a unified embodied system. AGIBOT’s portfolio spans humanoid robots, quadrupeds, dexterous systems, and commercial cleaning solutions. In March 2026, AGIBOT announced that its 10,000th robot had rolled off the production line.

Industry
Manufacturing & Production
Company Size
201-500 employees
Headquarters
Shanghai, CN
Year Founded
Unknown
Social Media