New year, new labels

Here at team homer we start the new year by spending hours labelling cups, cutlery and long-expired food items in pictures. Because we know how to have fun.

For those who aren’t familiar, labelling simply means to mark the region in an image where a specific object is. We do this for many pictures and then feed the labelled images into a neural network. This way, the network learns how to recognize the trained objects in new pictures even with perspectives not seen before.

We labelled more than 300 pictures of our many test objects, so Lisa and TIAGo should now be better equipped for recognizing and grabbing the right objects in the coming tests.

The whole group procrastin... concentrating on creating the perfect labelling set design.
The whole group procrastin... concentrating on creating the perfect labelling set design.
Dear diary, it's day 23 of labelling. I'm beginning to develop violent fantasies about smashing this hard-to-label translucent cup.
Dear diary, it's day 23 of labelling. I'm beginning to develop violent fantasies about smashing this hard-to-label translucent cup.
As service robots, Lisa and TIAGo can recognize all your cutlery (as long as your household only uses childproof plastic cutlery from a certain Swedish furniture concern).
As service robots, Lisa and TIAGo can recognize all your cutlery (as long as your household only uses childproof plastic cutlery from a certain Swedish furniture concern).

Nacht der Technik @ Handelskammer Koblenz

Last Saturday, many people got to meet Lisa and TIAGo at the “Nacht der Technik” (Technology Night) at Handwerkskammer Koblenz. The “Nacht der Technik” is an annual event organized by the chamber of crafts with the support of local tech companies. From noon until midnight, visitors can learn about latest technological advances in a number of small lectures and tutorials, live demonstrations and hands-on displays. Highlights of this year’s event included 3D-printing, autonomous driving and many AI and smart home related applications.

Lisa makes small talk with the crowd.
Lisa makes small talk with the crowd.

Our robots and our new student team supported our sponsor KEVAG Telekom at their booth. Although our robots still have difficulties understanding German, especially the younger guests were enchanted with Lisa. Plus, we had many interesting talks with visitors about our robots and related topics such as robots in geriatric care, robot bees and robot world domination. Thanks for dropping by!

This was also one of the first public events for our new student team which you will soon see more of – we’re currently preparing for our next challenge at the RoboCup German Open 2020 in Magdeburg.

TIAGo watches from behind and showcases pose recognition results on his screen.
TIAGo watches from behind and showcases pose recognition results on his screen.

Hat Trick

WE WON!!!!
WE WON!!!!

The preparation for todays finals meant a lot of stress for us last evening and also this morning. However, it went quite well in the end. The overall theme for the finals was food – so we came up with the “Lifecycle of Food”. That meant letting Lisa create a semantic map while doing an exploration of her (unknown) environment. So for example she knows where the food is stored.

Our award with the skyline of Sydney at Harbour Side
Our award with the skyline of Sydney at Harbour Side

In the meanwhile we showed how TIAGo can learn a task — like pouring some water into a cup — by demonstration. Just after showing us what he learned, we ordered TIAGo to clean the toilet like we did for last years World Robot Summit.

Ida is moving TIAGo's arm to train him how to pour water into a bowl
Ida is moving TIAGo's arm to train him how to pour water into a bowl

Competing against the Pumas from Mexiko in the finals, our performance – and also being already 1170 points ahead – lead us to the first place. Our friends from the new French team Catie achieved the third place. Conclusion: we made a hat-trick and are now the most successful ‘at-home’ team of the whole RoboCup. Yay!

Proud winners... Isabelle Kuhlmann, Patrik Schmidt, Christian Korbach, Mark Mints, Raphael Memmesheimer and Ida Germann
Proud winners... Isabelle Kuhlmann, Patrik Schmidt, Christian Korbach, Mark Mints, Raphael Memmesheimer and Ida Germann

Scores of the second day and we made it to the finals

This morning we participated in the restaurant game of stage 2 and had a very good run. Lisa reached the maingoal which was finding a calling customer, taking his order and serving it. She even managed to take the order of a second customer but the time for serving this whole order was a bit short. The difficulties in this task was that it took place in a real restaurant area which was completely unknown for the robot. By using depth mapping in combination with our standard navigation, Lisa successfully avoided all tables and chairs even if not detected by the laserscanner.

Lisa serving in a restaurant
Lisa serving in a restaurant
We also achieved quite some points in Where is This
We also achieved quite some points in Where is This
We kept leading the scores in stage two
We kept leading the scores in stage two