Robots in Everyday Human Environments: On Platform Development and Behaviour Dependent Control
Authors
Kracht, Simon ; Nielsen, Carsten
Term
10. term
Publication year
2007
Pages
130
Abstract
Dette speciale præsenterer udviklingen af en autonom mobil robot, der er designet til at fungere blandt mennesker. Hovedfokus er human-robot interaction (HRI), dvs. hvordan robotten opfører sig, når den møder og bevæger sig blandt mennesker. Vi introducerer en ny adfærdsalgoritme og validerer den både i simulation og i forsøg i den virkelige verden. Robotten tilpasser løbende sin adfærd til de personer, den møder, ved hjælp af case-based reasoning (CBR), en AI-tilgang der sammenligner nye situationer med tidligere eksempler lagret i en MySQL-database. Ud over at interagere med mennesker kan robotten bestemme sin position (lokalisere sig), navigere i omgivelserne og undgå uventede forhindringer. Al softwaren, inklusive adfærdsalgoritmen, er udviklet til Player-rammeværket for robotsoftware og implementeret på en FESTO Robotino. Da Player-rammeværket tilbyder mange robotfunktioner, kan systemet danne grundlag for fremtidig robotforskning. Eksperimenter på robotten viser lovende resultater for både adfærden og systemet som helhed.
This thesis presents the development of an autonomous mobile robot designed to operate around people. The main focus is human-robot interaction (HRI), meaning how the robot behaves when it meets and moves among humans. We introduce a new behavior algorithm and validate it in both simulation and real-world experiments. The robot continually adapts its behavior to the people it encounters using case-based reasoning (CBR), an AI approach that compares new situations to previous examples stored in a MySQL database. Beyond interacting with people, the robot can determine its own position (localize), navigate through its environment, and avoid unexpected obstacles. All software, including the behavior algorithm, is built for the Player robot software framework and deployed on a FESTO Robotino. Because the Player framework supports many robot capabilities, the system can serve as a foundation for future robotics research. Experiments on the robot show promising results for both its behaviors and the overall system.
[This abstract was generated with the help of AI]
Documents