Building Intelligent Robotics Systems with Distributed Components

Abstract

Building an intelligent robot system has been an extensive research area. There are many advances in components needed to construct the robotic system, such as vision systems, sensory systems, planning systems, among others. Integration of this components represents a big challenge for robot designers, due to they come from different vendors and with different interfaces or operating systems. This is more difficult if the overall system development has to deal with environmental uncertainties or changing conditions. In these cases, new tools and equipment are necessary to adapt the initial configuration to the new changing requirements. Each added component increases the complexity of the system due to the interconnection required with the previous components.
In this work, we present an approach to solve this integration problem using concepts of distributed computing areas. We named this concept Wrapper Components. This concept is based on a standard middleware software specification. Wrapper components are object-oriented modules that create an abstract interface for a specific class of hardware or software components. If these components provide "intelligent" functions, the overall system is capable of show some basic smart behavior through specific actions to react under changes in the environment. We tested our approach by solving an experimental classical problem named block-world. The intelligent functions are object recognition, environment recognition, planning, tracking capabilities and robot arm control.