Mowing lawns is boring and repetitive work. Existing ('autonomous') systems generally use wiring to detect the end of the lawn so the mower knows when to stop and turn. GPS accuracy is too low to fix the dependency on the wiring.
We think we can do better by building a 'smart' autonomous lawn mower, utilizing deep learning vision systems that are now emerging. The repetitiveness of the lawn is exactly why we think the mowing robot can be improved and the tapas board fits perfectly for the required control of the electronics involved.
Hardware set up:
Raspberry Pi as the motherboard + 2 camera PI modules (stereo vision)
2 Tapas board for drive-system (tankstyle steering, 2 brushless motors)
1 Tapas board for controlling the mower blades (powered by 1 brushless motors, initially, see below)
We have some ideas to improve the mowing blades (sort of a gimbal/height adjustment system based on the surroundings), but we are keeping that for a later version.
Our general background and education is in software development, but in our daily jobs we have designed software, hardware and 3D printed enclosures and parts for IOT developments.
Niels Brouwers: Software development and hardware design
Peter Schennink: Software development and hardware design
Hanna Husken: Software development and 3D printing