• Train a Robot - Part 2

Christopher Stumm
"Robotizing data centers" – the key to this project’s success are the partners we’re working with.

A well-known African proverb says that it takes an entire village to raise a child. Similarly, it takes a whole lot of know-how to train a robot! We’re delighted to have secured for our project 14 partners who are actively involved – and enthusiastic about it. This was very evident during our first workshop at the Global NTT Innovation Lab in Frankfurt. Getting to know each other was the first item on the agenda. We wanted answers to questions like: Who has what experience? How can we best pool our collected expertise? In order to ensure that our work would be even more purposeful in the future, together we defined six focus areas. Partners could participate in one or more categories, as shown in the following overview:​

As far as the robot itself is concerned, the next item on the agenda is visual training. Our plan is to mount several cameras on the robot so that it can "see". Once this step has been taken, we’ll gather as much image and video data as possible to serve as a database for pattern recognition. Then we’ll train the artificial intelligence models on these patterns – but you’ll have to wait for our next video to learn more about it...