Simulation-Based Robot Placement Using a Data Farming Approach

Abstract

Increasing flexibility in production systems is driving the use of robotic solutions. During their planning, robots must be placed according to their future operations. Thereby, influences such as space limitation, mechanical reach or cycle time must be taken into account. This paper introduces a concept based on the data farming methodology aiming at the optimal robot positioning for a given set of constraints. By simulating a defined sequence of robot operations with changing robot placement in a definable investigation area, each result data set is stored and analyzed. The simulation run with the best fitting robot position according to the defined key performance indicators is shown. For further evaluation, a clustering algorithm is used to evaluate the simulation results. The usage of the proposed method enables production planners to conveniently place robots in the optimal position according to their later application.