WedAI - autonomous cleaning robot
| Reference number | |
| Coordinator | Weda AB |
| Funding from Vinnova | SEK 500 000 |
| Project duration | December 2020 - August 2021 |
| Status | Completed |
| Venture | AI - Competence, ability and application |
| Call | Start your AI-journey for public organizations - autumn 2020 |
Important results from the project
The goal of the project was to transform an already existing underwater pool cleaning robot into an AI driven robot, through 2 specially developed AI functions: 1-Path planning: the robot is fed with the pool map and can design the optimized path to clean it. 2-Object recognition: a camera is installed on the cleaning robot and the AI can identify possible objects on the bottom of the pool. Both functions now work outside of the robot but the installation is still not satisfactory.
Expected long term effects
The AI functions run well and show new future development possibilities: integration of a sonar like system sending the actual position as a feed back. Object recognition can also be further developed and tailored to different customers through specific databases. As of today, the results are not fulfilling the expectations; the robot is still not piloted by the AI functions. The installation still requires 2 meaningful improvements: the cleaning robot’s turn needs to be reprogrammed and the computer board needs to be upgraded to a more performant model.
Approach and implementation
The implementation or the installation in the robot is the weak point of the project. The robot can only turn while moving forward or rearward, since it is programmed like this by its PLC; there could be a mechanical limitation as well, still the PLC reprogramming should be the way to fix this issue. With this change made, it is possible to continue the implementation and testing of the path panning function. The computer board performance limitations clearly prevented the team from gaining more experience with the object recognition function.