So far, most AIs in robotics have been used to recognise, grip and precisely position objects.
Using the object to fulfil a certain purpose in a changing environment has not yet been sufficiently researched.
In order to optimise the software in our research project and to identify and fix errors in good time, the system was regularly trained and tested during the development phase.
With a mix of real hardware and simulation, the system could be tested under different conditions. Thanks to error detection, our algorithms could be adapted to achieve even better results and to fix the errors found.
Here you will find all articles about our research project.
Image credit: AI-generated with Canva Pro
kununu