← Back to list
Robot AI Model & Algorithm Engineer
A Physical AI Company
Responsibilities
- Lead the R&D and implementation of intelligent perception and behavioral decision-making algorithms for robots, including but not limited to Reinforcement Learning (RL), Imitation Learning (IL), and Vision-Language-Action (VLA) models, enhancing generalization and complex task execution in real-world environments.
- Design and optimize multi-modal learning architectures that fuse visual, linguistic, and action data to drive robot understanding and interaction capabilities in dynamic, real-world scenarios.
- Build system-level algorithm integration frameworks based on ROS / ROS2, ensuring real-time performance (e.g., response < 100ms) and stability to support whole-machine scheduling and closed-loop control.
- Deploy low-level control algorithms such as motion control, path planning, and force control onto physical robot platforms, achieving efficient and precise operations.
- Research and implement cutting-edge learning frameworks like VLA and RL / IL for robotics, driving the transition from simulation validation to real-system deployment.
- Establish standardized processes for data processing, model training, and evaluation to continuously improve model performance and robustness.
- Write technical proposals, design documents, and research reports; contribute to team capability building with opportunities to publish high-quality academic papers or technical articles.
Requirements
- Master’s degree or above in Computer Science, Artificial Intelligence, Control Science and Engineering, Robotics, or a related field.
- Solid foundation in machine learning and robotics, with familiarity in kinematics, dynamics, control theory, SLAM, and motion planning.
- Proficient in C++ and Python, with strong software engineering skills and system-level debugging capabilities.
- Familiar with ROS / ROS2 development and knowledgeable about robot system architecture and debugging workflows.
- Practical experience in reinforcement learning, imitation learning, and multi-modal learning (e.g., behavior cloning, policy gradients, VLA models).
- Prior experience in robot system deployment is preferred (including simulation validation and real-robot joint tuning).
- Strong teamwork skills, problem-solving and analysis abilities, and a habit of thorough documentation.
About the company
The core founding team includes Professor Gao Shenghua, an expert in computer vision, Professor Yang Yanchao, a robotics specialist, and other senior experts in artificial intelligence and industrial engineering. The team combines cutting-edge research capabilities with extensive experience in complex system deployment. Joyang Consulting focuses on intelligent memory and autonomous learning, specializing in visual perception and motion control. Our goal is to build an 'intelligent brain' that understands the physical world, endowing robots with human-like perception, reasoning, and action capabilities.
- Industry
- AI Applications & LLMs
- Founded
- 2023-09-20