DocumentCode :
695122
Title :
Inferring guidance information in cooperative human-robot tasks
Author :
Berger, Erik ; Vogt, David ; Haji-Ghassemi, Nooshin ; Jung, Bernhard ; Ben Amor, Heni
Author_Institution :
Inst. of Comput. Sci., Tech. Univ. Bergakad. Freiberg, Freiberg, Germany
fYear :
2013
fDate :
15-17 Oct. 2013
Firstpage :
124
Lastpage :
129
Abstract :
In many cooperative tasks between a human and a robotic assistant, the human guides the robot by exerting forces, either through direct physical interaction or indirectly via a jointly manipulated object. These physical forces perturb the robot´s behavior execution and need to be compensated for in order to successfully complete such tasks. Typically, this problem is tackled by means of special purpose force sensors which are, however, not available on many robotic platforms. In contrast, we propose a machine learning approach based on sensor data, such as accelerometer and pressure sensor information. In the training phase, a statistical model of behavior execution is learned that combines Gaussian Process Regression with a novel periodic kernel. During behavior execution, predictions from the statistical model are continuously compared with stability parameters derived from current sensor readings. Differences between predicted and measured values exceeding the variance of the statistical model are interpreted as guidance information and used to adapt the robot´s behavior. Several examples of cooperative tasks between a human and a humanoid NAO robot demonstrate the feasibility of our approach.
Keywords :
Gaussian processes; accelerometers; human-robot interaction; humanoid robots; intelligent robots; learning (artificial intelligence); manipulators; pressure sensors; regression analysis; Gaussian process regression; accelerometer; behavior execution; cooperative human-robot tasks; direct physical interaction; force sensors; guidance information; humanoid NAO robot; jointly manipulated object; machine learning approach; periodic kernel; pressure sensor information; robotic assistant; sensor data; statistical model; training phase; Kernel; Legged locomotion; Predictive models; Robot kinematics; Robot sensing systems; Stability analysis;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Humanoid Robots (Humanoids), 2013 13th IEEE-RAS International Conference on
Conference_Location :
Atlanta, GA
ISSN :
2164-0572
Print_ISBN :
978-1-4799-2617-6
Type :
conf
DOI :
10.1109/HUMANOIDS.2013.7029966
Filename :
7029966
Link To Document :
بازگشت