Brand Overview
LimX Dynamics
Bipedal locomotion platforms for the embodied-AI decade.
- 2022Founded
- Shenzhen, ChinaHeadquarters
- Bipedal and humanoid robotsSpecialty
- 2 research platformsAt RoboticsSelect
LimX Dynamics is a Shenzhen-based legged-robotics company founded in January 2022. It develops bipedal and humanoid platforms aimed squarely at researchers, universities, and developers working on locomotion control and embodied AI. RoboticsSelect carries its two education-focused platforms: the TRON1 EDU multi-modal biped and the Oli EDU humanoid.
What makes LimX Dynamics different
LimX Dynamics was founded in 2022 in Shenzhen and concentrates on one hard problem: making legged robots walk reliably in the real world. Rather than chasing a single flagship humanoid, LimX builds a family of locomotion platforms and publishes the tooling researchers actually need - open SDKs in C++ and Python, ROS 2 packages, calibrated simulation models for MuJoCo and Isaac Sim, and reinforcement-learning training examples with sim-to-real transfer pipelines on GitHub.
The TRON1 is the approach in miniature: LimX markets it as the first multi-modal biped, a platform built to study bipedal locomotion, balance, and policy deployment on real hardware. The Oli EDU humanoid extends the same control philosophy and SDK family to humanoid research and developer education, so labs can prototype in simulation and deploy to hardware without switching ecosystems. For university groups moving from simulated agents to physical robots, that continuity is the whole point.
The LimX Dynamics lineup at RoboticsSelect
We carry both current LimX research platforms: the TRON1 EDU multi-modal biped for locomotion and reinforcement-learning research, and the Oli EDU humanoid for embodied-AI coursework and development. Both are research instruments rather than consumer products - if you are specifying one for a lab, our team can help with configuration and institutional ordering.
Research platforms
Both platforms expose control from high-level locomotion commands down to joint level, and both slot into the standard research toolchain: develop and train in simulation against the published models, validate policies, then deploy the same code to hardware. Typical buyers are robotics and control labs, embodied-AI groups, and engineering departments building a legged-robotics track into their curriculum.