Incremental Learning of Tasks From User Demonstrations, Past Experiences, and Vocal Comments

Abstract

Since many years the robotics community is envisioning robot assistants sharing the same environment with humans. It became obvious that they have to interact with humans and should adapt to individual user needs. Especially the high variety of tasks robot assistants will be facing requires a highly adaptive and user-friendly programming interface. One possible solution to this programming problem is the learning-by-demonstration paradigm, where the robot is supposed to observe the execution of a task, acquire task knowledge, and reproduce it. In this paper, a system to record, interpret, and reason over demonstrations of household tasks is presented. The focus is on the model-based representation of manipulation tasks, which serves as a basis for incremental reasoning over the acquired task knowledge. The aim of the reasoning is to condense and interconnect the data, resulting in more general task knowledge. A measure for the assessment of information content of task features is introduced. This measure for the relevance of certain features relies both on general background knowledge as well as task-specific knowledge gathered from the user demonstrations. Beside the autonomous information estimation of features, speech comments during the execution, pointing out the relevance of features are considered as well. The results of the incremental growth of the task knowledge when more task demonstrations become available and their fusion with relevance information gained from speech comments is demonstrated within the task of laying a table.

Robot training using system identification

“…Programming mobile robots by demonstration is now a major trend in the robotics community [Pardowitz et al, 2007,Allissandrakis et al, 2005. Many researchers demonstrate the viability of this approach in tasks such as maze navigation [Hayes and Demiris, 1994] [ Demiris and Hayes, 1996], arm movement [Schaal, 1997] [ Calinon and Billard, 2007] or service robotics [Demiris and Johnson, 2003].…”

Intention deduction from demonstrated trajectory for tool-handling task

“…Dautenhahn and Nehaniv (2002) proposed an approach for the robot to learn from human demonstration by imitation, referred to as the correspondence problem, and later the team developed a system that can learn two-dimensional (2D) arranging tasks (Alissandrakis et al 2005a,b). Pardowitz et al (2006) proposed a hierarchical structure for the robot to deal with complex tasks while the motion order can be changed, and later they went on analyzing human motion features for high-level tasks (Pardowitz et al 2007). With both symbolic and trajectory levels of skill representation, Ogawara et al (2003) proposed a method that determines the essential motions from the possible motions.…”

Hierarchical Bayesian Modeling of Manipulation Sequences from Bimodal Input

“…In general, one is interested in autonomous identification of action primitives in the context of imitation learning and human-machine interaction (Sanmohan, Krüger, & Kragic, 2010;Takano & Nakamura, 2006). Within this domain, Matsuo et al focused on force feedback (Matsuo, Murakami, Hasegawa, Tahara, & Ryo, 2009) while a combination of different sensors like CyberGlove, Vicon or magnetic markers and tactile sensors has been used by (Pardowitz, Knoop, Dillmann, & Zöllner, 2007), (Kawasaki et al, 2000) and (Li, Kulkarni, & Prabhakaran, 2006). In (Zöllner, Asfour, & Dillmann, 2004) a bimanual approach is described.…”

“…Our method does not employ any specific knowledge about the components of the action sequence. Based on two simple models, the modeling does not require a large set of domain-specific heuristics describing each action primitive as is commonly the case in similar approaches (Pardowitz et al, 2007;Kawasaki et al, 2000;). Due to the simplicity of these two fundamental models and the modeling concepts used within our approach, the developed procedure can be easily used in a wide range of scenarios, like imitation learning, cooperation and assistance.…”