// the find
mithi/hexapod-robot-simulator
A hexapod robot simulator built from first principles
A browser-based hexapod robot simulator that implements forward and inverse kinematics from scratch using only NumPy for math and Plotly Dash for the UI. It's a learning project, not a production tool — the author has already rewritten it in JavaScript and says that version is orders of magnitude faster. The target audience is robotics students and hobbyists who want to see how hexapod IK works without wading through ROS.
The kinematics are implemented from first principles with minimal dependencies — just NumPy — which makes the math readable and educational rather than hidden behind a framework. The UI is genuinely interactive: you can manipulate body dimensions, joint angles, and camera view in real time, which is useful for building intuition. There's a physical robot tie-in (hexapod-irl fork) showing the IK solver was validated against actual hardware, not just simulation. Test cases are organized by scenario with expected pose outputs, which is the right way to test geometric code.
Performance is bad enough that the author calls it out themselves and recommends switching to the JavaScript rewrite — Plotly Dash re-renders the full 3D scene on every slider move, which kills interactivity on anything complex. The stability check (will the robot fall over?) is marked WIP and never landed, so the simulator can produce physically impossible poses without warning. Python 3.8.1 and Dash 1.18.1 are pinned from 2021 and the project hasn't had meaningful commits in years, so dependency rot is a real issue if you try to run it today. Test coverage exists but is thin — a handful of hand-coded cases, no property-based testing of the IK solver's edge cases like near-singular configurations.