Warsaw RAPP Hackathon 2016

Warsaw RAPP Hackathon 2016

Local edition of RAPP Hackathon was held in Warsaw on May 20-21, 2016. The first evening was devoted to an introductory lecture and tutorials. Prof. Cezary Zieliński delivered a lecture about the ideas on which the RAPP project is based and described the overall structure of the RAPP system. After that Maciej Stefańczyk and Wojciech Dudek presented the capabilities of the robots, especially focusing on navigation, vision and communication modules. The RAPP Platform services were also presented as an extension to robot skills. The day ended with a short brain storming session on possible applications that can be created using the RAPP infrastructure.

The next day was fully devoted to application development. Eight teams (each consisting of two people) had 10 hours to implement at least a part of their planned RAPP Application. There were two robots prepared to work within the RAPP infrastructure: NAO humanoid and Elektron mobile robot. To make the whole development and testing process quicker, simulator for the Elektron mobile robot had been prepared. It allowed the participants to test their Rapps without the real robot.

After ten hours of development, each team had a few minutes to present their idea and what they managed to implement. Half of the applications were prepared for the NAO robot, another half targeted the mobile robot. Below a brief description of the applications using the RAPP infrastructure and diverse capabilities of the robots are presented. Those applications were chosen in such a way that they would act as examples of Rapps for the elderly users.

Drug Assistant. Application has two modules – reminder and helper. The reminder module keeps track of medicine doses and treatment time and, at a prescribed time, asks the user whether he/she already had taken them. The helper module assists the user with selecting the proper box of medicines. When the user has problems with recognition of the proper medicine box or forgot which drug to take at the moment, he or she can show the robot all the medicines and the robot, by using platform services for recognition of known objects, points to the proper box with its hand and reminds the treatment dose. The robot tries to keep eye-contact with the user when talking.

Used hardware and software services:
Robot: NAO
Robot modules: communication (text to speech, word spotting), navigation (arm gestures, head movement), vision (image capture)
Platform modules: object recognition, face detection

Picking up small objects dropped on the floor. Elders often have problems with picking things dropped on the floor. The robot walks around the room and when it spots something lying on the floor recognizes its characteristics (such as color, size) and asks the person, if he or she wants to pick it up. If so, the robot tries to pick up the object and raise it high enough for the user to grasp it.

Used hardware and software services:
Robot: NAO
Robot modules: communication (text to speech, word spotting), navigation (walking, assuming a predefined pose, arm movement, head movement), vision (image capture)
Platform modules: path planning

Navigation assistant in public spaces. Older people sometimes feel lost in new places, especially when they have to find a particular room or person (e.g. someone’s office). They have problems with finding the way around and with reading labels on the doors. The robot can be used as an assistant, that asks people where they want to go and then either follows the path planned (if target position is known) or assists the person and recognizes and reads the text written on labels if the person needs it. In both scenarios the robot is constantly tracking the position of the human to adjust its speed (not to move too slow or too fast).

Used hardware and software services:
Robot: Mobile base with vision sensors mounted on a high-platform fixed to the base (Elektron)
Robot modules: communication (text to speech, word spotting), navigation (movement, head movement), vision (image capture)
Platform modules: path planning, face detection, posture detection

IMG_5012IMG_5007IMG_4983IMG_5001IMG_5009 IMG_4980IMG_4974IMG_4958IMG_4927IMG_4915IMG_4939 IMG_4949IMG_4953IMG_4951