Control a iM.Master toy robot from Python instead of its official phone app.
Build a simple see-think-act loop that lets a local AI model steer the robot using camera input.
Test robot control logic on a laptop without hardware using the built in dry run mode.
Study a real example of reverse engineered Bluetooth Low Energy protocol decoding.
| 2alf/im.master-sdk | 0marildo/imago | 0xdfi/glm-5.2-1m-4x-dgx-spark | |
|---|---|---|---|
| Stars | 3 | 3 | 3 |
| Language | Python | Python | Python |
| Setup difficulty | moderate | easy | hard |
| Complexity | 3/5 | 2/5 | 5/5 |
| Audience | developer | general | researcher |
Figures from each repo's GitHub metadata at analysis time.
Controlling the physical robot needs a Raspberry Pi running as root, the AI addon also needs a local LLM runtime like Ollama.
iM.Master SDK is a Python library for controlling a toy robot called iM.Master that normally only works through its own phone app. The robot has no official remote control or programming interface, so the project figured out how the app talks to the robot over Bluetooth Low Energy and rebuilt that communication in plain Python, letting you drive the robot directly from code with commands like moving forward for a set number of seconds or spinning in place. The core of the SDK has no external dependencies and only needs the Python standard library. On top of that, there is an optional add-on that lets a small AI model steer the robot. You can point a camera, such as an old phone running a camera app, at the robot's surroundings, run object detection on a Raspberry Pi, and have a local language model decide which direction to move based on what the camera sees. The project describes running this whole loop, camera to detector to AI to robot, entirely on local hardware with no cloud service involved. Because moving the actual robot requires low level Bluetooth access, that part only works on Linux, in practice a Raspberry Pi running the script as an administrator. The pure command building and decoding logic, however, works on any operating system, including Windows and macOS, and there is even a simulated robot mode that lets you test the whole AI control loop without any physical hardware connected. The SDK includes a safety feature: if the control program crashes or stops sending commands, the robot automatically stops itself instead of continuing to drive blindly. The project is released under the MIT license, is written in Python 3.10 or newer, and can be installed either through pip or by cloning the repository directly.
A Python SDK that lets you drive a toy robot called iM.Master from code, with an optional local AI vision addon.
Mainly Python. The stack also includes Python, Bluetooth Low Energy, Raspberry Pi.
MIT license, free to use, modify, and share as long as the original copyright notice is kept.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Don't trust strangers blindly. Verify against the repo.