Modelling, Specification and Robustness Issues for Robotic Manipulation Tasks

In this paper, a system for modeling of service robot tasks is presented. Our work is motivated by the idea that a robotic task may be represented as a set of tractable modules each responsible for a certain part of the task. For general fetch-and-carry robotic applications, there will be varying demands for precision and degrees of freedom involved depending on complexity of the individual module. The particular research problem considered here is the development of a system that supports simple design of complex tasks from a set of basic primitives. The three system levels considered are: i) task graph generation which allows the user to easily design or model a task, ii) task graph execution which executes the task graph, and iii) at the lowest level, the specification and development of primitives required for general fetch-and-carry robotic applications. In terms of robustness, we believe that one way of increasing the robustness of the whole system is by increasing the robustness of individual modules. In particular, we consider a number of different parameters that effect the performance of a model-based tracking system. Parameters such as color channels, feature detection, validation gates, outliers rejection and feature selection are considered here and their affect to the overall system performance is discussed. Experimental evaluation shows how some of these parameters can successfully be evaluated (learned) on-line and consequently improve the performance of the system.