Open robotics organization Est. 2026
Mechatronics, Intelligence, Actuation & Operation
Robotics from first principles to real-world impact. We model motion, learn policies, build hardware, and put machines to work.
Robotics Embodied intelligence From motion to real-world impact
01 Four disciplines
miao brings the physics, the learning, the hardware and the field into one conversation. Each pillar is useless alone; together they close the distance between a model and a working machine.
Model · Control · Motion
Contact · Simulation
Before a robot moves, we model how it can. Kinematics, rigid-body dynamics, contact and control connect every policy to the physics of the machine, in simulation and in the real world.
Perception · Learning · Reasoning
Decision · Foundation Models
Sensing becomes decisions. We train policies in simulation, ground them in real observations, and test whether they generalise beyond the conditions they were born in.
Actuators · Mechanisms · Hardware
Power · Sensing
Intent becomes torque. Motors, transmissions and electronics decide whether a controller survives contact, so we treat the hardware as part of the algorithm, not a detail after it.
Locomotion · Manipulation
Task Execution · Deployment
Locomotion, manipulation and task execution in unstructured space. Field deployment is the experiment no simulation can replace, and the place where the next question comes from.
02 How it fits together
Models meet hardware, hardware meets the world, and the world sends data back. Every deployment is also a measurement.
Dynamics, contact and control define what the machine can physically do.
MechatronicsSimulation turns that model into policies, and policies into behaviour.
IntelligenceMotors and mechanisms turn commands into motion with limits of their own.
ActuationReal terrain, real payloads, real failure modes. The field sets the next problem.
Operation03 Organization projects
Projects in miao-org — tools, research resources and open community experiments.
A local dashboard for robotics paper digests and updates from researchers, labs and companies.
Organizes chat messages into tasks, events and linked topics in a local dashboard, with optional AI assistance.
A curated guide to SuperMimic implementations, extensions and demos, with links to the original work and learning resources.
A bilingual Markdown framework for a robotics engineer learning plan, published with Docsify.
A blank repository and a standing invitation: open a PR and show us what robotics software can become.
Explore all repositories in miao-org on GitHub.
04 Open by default
We publish code, share what works, and keep the experiments visible. If you build robots, you already have something to add.