Automated Analysis of Assembly Processes in Human-Robot Collaboration: Research Approaches and Challenges

Abstract

Due to the increasing trends of individualization and variant diversity, cobotic production systems are becoming more important. However, cobots are hardly used in small and medium-sized enterprises (SME) because of a lack in expertise, experience or resources. In order to be able to assess whether the use of a cobot is reasonable, there are already many methods in research with different focuses such as ergonomics, productivity or economic efficiency. However, many methods are not targeted for use by SME, thus still require the involvement of experts, and often only consider economic efficiency without consideration of human-centered aspects. Furthermore, these are mostly not integrated into commonly used process planning tools.This work provides an overview of methods for automated analysis of assembly processes in HRC. We present a method for HRC process analysis with an individual weighting of fatigue, time, costs and safety that can be set depending on an application-specific focus defined by the user. It allows non-experts to perform a process analysis based on the individual weighting and a digital twin of the workstation. An evaluation method developed for this purpose calculates the benefit of a cobot in the analyzed process depending on the preset focus. To keep the effort low, the method includes a workflow to easily create a digital twin of the workplace with sufficient accuracy.In addition to the pre-estimation of the human-cobot process, the method allows an easy planning of collaborative production processes, which is why we suggest a conceptual integration into a process planning and control framework. The proposed approach is demonstrated in an use case by analyzing the potential for introduction of collaborative robots for the production of currently manually assembled solar inverters.