Maintenance Task Classification: Towards Automated Robotic Maintenance for Industry

Abstract

The business model of high-value capital assets is shifting from purchasing a physical product to acquiring a result or a function supported by the product combined with a number of related services. One such service, maintenance, is perhaps the most efficient way to keep the function available during the product lifecycle. Automation has played a vital role in industry throughout history, particularly within the production line. With the movement towards providing product service systems the need for services such as maintenance are increasingly important for a manufactured product, and the pull towards automation may drive down costs and improve performance time. Although currently robotic applications to maintenance beyond monitoring and inspection tasks are not common, this research aims at exploring the feasibility of future maintenance robots that can perform a variety of maintenance tasks. As its first step, this work looks first at investigation, cataloging and classification of a number of maintenance tasks using standard industrial engineering techniques such as time motion, method or workflow analysis. This involves decomposing the maintenance work into a number of ‘unit tasks’ required to be performed in order to accomplish the specified maintenance.