Python clean algorithm (highly experimental) for Neato vacuum cleaner robot and built-in simulator. If you have have interface details (serial USB etc.) for another vaccum cleaner robot, you can use the clean algorithm code with it too.
The robot is equipped with two gear motors and with 2D LiDAR sensor - this is the only sensor used in this code. The LiDAR sensor is used for mapping & localization (aka 'SLAM') and obstacle detection.
The robot can be controlled via serial USB interface from this Python software:
Simulator session run (a simulator for running the code is built-in - click on 'play' button for playback):
- Phase 'exploring': Using only the LiDAR sensor, the robot follows the walls until it gets back to its start position.
- Phase 'cleaning': The robot looks in the local area for free cells as targets. If there are no targets in the local area, it performs a global search for a free cell (using 'astar algorithm').
Type this into your Python3 environment (tested only on Linux Anaconda Python3.7 environment):
- git clone https://github.com/owlRobotics-GmbH/vacuum_cleaner
- pip install -r requirements.txt
- python cleaner.py --sim (for simulator)
- python cleaner.py (for Neato robot connected via serial USB)
- Try out the code and tune it:
- Add obstacle detection during cleaning, mark the cells in front of the robot as obstacles, and trigger the find route to next free cell code, e.g. also consider these obstacle cells in the astar-algorithm for finding the next free cell.
- Build your own robot with owlRobotics platform (https://github.com/owlRobotics-GmbH/owlRobotPlatform)
- Start your carreer at owlRobotics (https://owlrobotics.de/index.php/en/home-of-robotic-solutions/careers)
- ...