SSRobotics · Source Code Collection
Part of Humanoid Robot Hub · 🔎 Search all resources
IK, inverse dynamics, MPC / DDP, motion planning, optimization and reinforcement learning.
IK, Inverse Dynamics, MPC / DDP, Motion Planning, Optimization และ Reinforcement Learning
Curated by SSRobotic — Robotics Engineer & Open-source Curator. This repository contains actual source files and models in projects/, alongside a resource catalog and engineering guide. The imported implementations are unchanged snapshots credited to their original creators.
898 tracked upstream files — เปิดโฟลเดอร์ด้านล่างเพื่อดูโค้ดจริง โมเดล และตัวอย่างต้นฉบับ
| Project files | Files | Pinned source | License / notices |
|---|---|---|---|
| pink-kinematics/pink | 128 | 2c333161fc26 |
LICENSE · NOTICE |
| loco-3d/crocoddyl | 770 | 0d3b1ae9c2f0 |
LICENSE |
Source manifest records commits, file counts, Git LFS pointers and external submodules. Per-file manifests record original Git blob hashes. projects/ keeps upstream code, README files, licenses and notices; SSRobotics supplies the organization and guides.
git clone --depth 1 https://github.com/SSRobotic/humanoid-algorithms.git
cd humanoid-algorithmsImplementation · Examples · Behavior tests
python -m pip install -r requirements.txt
python examples/run_ik.py
python examples/run_mpc.py
python examples/run_rrt.py
python -m unittest discover -s tests -vThe original SSRobotics examples implement damped least-squares IK for a planar two-link arm, unconstrained finite-horizon linear MPC for a 1D double integrator, and collision-checked RRT for a 2D point robot. They require only NumPy and run on CPU. They are learning examples, with no humanoid balance or hardware integration. Pink and Crocoddyl provide the imported robotics implementations; install their dependencies using their retained README files.
Crocoddyl’s pinned cmake submodule is listed in its manifest and .gitmodules, and is not bundled as a nested Git repository. Use the original upstream checkout with --recurse-submodules for source builds that need it.
Use the algorithm map and engineering guide to choose between IK, inverse dynamics, MPC / DDP, planning, numerical optimization and RL.
| Resource | What it provides | Keywords | Original source |
|---|---|---|---|
| Pinocchio | Rigid-body dynamics and kinematics algorithms for robotics, including model-based control research. | Kinematics, Dynamics, C++ | stack-of-tasks |
| Drake | Model-based robotics toolbox for dynamics, optimization, planning and control. | Dynamics, Optimization, Planning | RobotLocomotion |
| RSL RL | Learning-algorithm implementations used in robotics training workflows. | RL, Training, Robotics | leggedrobotics |
| MoveIt 2 | ROS 2 motion-planning and manipulation framework. | ROS 2, Motion planning, Manipulation | moveit |
| TSID | Task-space inverse-dynamics control using Pinocchio, with robotics exercises and examples. | Inverse dynamics, Whole-body control, QP | stack-of-tasks |
| Crocoddyl | Contact-aware optimal control with DDP-family solvers for robot motion. | DDP, Optimal control, Contacts | loco-3d |
| OCS2 | Optimal-control and MPC tooling for systems with switching dynamics. | MPC, Optimal control, Switched systems | leggedrobotics |
| Pink Inverse Kinematics | Differential inverse kinematics in Python using Pinocchio and QP solvers. | IK, Pinocchio, QP | pink-kinematics |
| OMPL | Sampling-based motion-planning algorithms; robot collision checking and execution are supplied by integrations. | Motion planning, RRT, Sampling | ompl |
| cuRobo | GPU-accelerated robot motion-generation tools; verify the API version and supported robot setup upstream. | CUDA, Trajectory optimization, Motion planning | NVlabs |
| OSQP | Numerical solver for convex quadratic programs, useful as a component in IK and control formulations. | QP, Optimization, Solver | osqp |
| CasADi | Symbolic differentiation and optimization tools for formulating nonlinear control and trajectory problems. | NLP, Automatic differentiation, MPC | casadi |
| Stable-Baselines3 | PyTorch reinforcement-learning implementations, including PPO, SAC and TD3; supply a compatible task environment. | RL, PPO, SAC | DLR-RM |
resources.json contains the structured entries. GUIDE.md connects tools to tasks and records the information needed to reproduce an experiment.
🌐 Browse all 14 categories
🦾 Models & URDF · 🌐 Simulation · 🧠 Robot Learning · ⚙️ Whole-body Control · 👁️ Vision & 3D Perception · 🤏 Manipulation & Planning · 🎮 Teleoperation · 💾 Datasets & Demonstrations · 💬 Vision-Language-Action · 🔌 Hardware & SDKs · 📡 ROS & Integration · 📚 Research & Benchmarks · 🟢 Isaac Sim & Isaac Lab · 🧮 Robotics Algorithms
⭐ Star the Hub · 👤 Follow SSRobotic · 💡 Suggest a resource
Suggestions and corrections follow the Hub contribution guide. Linked software, assets and documentation keep their original licenses. The original SSRobotics logo is supplied by its creator; the root MIT license covers our original descriptions, guides and SSRobotics example code only. Imported projects/ files retain their own licenses.