TRIZ Driven Identification of AI Application to Improve Navigation of Mobile Autonomous Robots

Abstract

Increase of effectiveness of navigation in the case of autonomous mobile robots with limited sensory systems integrated that operate in different types of production facilities is considered in this research work. The challenge is to define a reliable solution in a cost-effective design. The research methodology integrates TRIZ within the framework of voice-of-use-table-performance function deployment (VOUT-PFD) design planning framework. The key user requirements and engineering specifications defined with VOUT-PFD have been analyzed in terms of correlations. For the identified sets of negative correlations, TRIZ Contradiction Matrix has been considered to formulate the generic areas for inventive problem-solving. Using the method of weighted analysis of interdependencies (AIDA), the compatible TRIZ vectors have been selected for guiding the design of the artificial intelligence (AI) algorithm. These vectors have been introduced in the framework of Complex System Design Technique (CSDT) in relation with generic modules of the AI system (algorithm, related inputs from the sensors and mechanical limitations of the robotic system) in order to design the navigation solution. In this article eight areas of possible improvements using AI algorithms have been found and the first area of research, which represents the robot construction is further detailed. The major result of this paper is that it shows a structured way in which inventive problem-solving thinking can lead to possible improvement areas regarding navigation of autonomous mobile robots (AMRs) in industrial environment.