Dynamic task sharing strategy for adaptive human-robot collaborative workcell

Abstract

Industry 4.0 prompts the re-organization of production systems with flexible collaborative workcells. Half way between the manual and the full automated cell, the new cell is configured by having humans and robots working together and collaborating to the execution of complex working tasks. Considering the production process subdivided in work tasks, it is apparent that some tasks could be performed more proficiently by humans or robots alone, others collaboratively. Analogy can be found with the job assignment problem in conventional production, for which a wide range of strategies and procedures has been researched. In Human-Robot Collaboration the procedures should be redefined having new objectives and following different rules. This is the intent of the study: to classify which tasks should be executed by men, robots or by both collaboratively. The classification method makes use of a training set of pre-classified tasks. Then task assignment combined with job scheduling is introduced to avoid resource overload.