Rapid deployment of remote laser welding processes in automotive assembly systems
CHALLENGE - Remote laser welding (RLW) has received increased attention in recent years due to its benefits of processing speed, lower investment, cost per stitch, and process flexibility. Its potential in automotive assembly remains underexploited due to challenges involving system, process and fixture design, and part variation. An integrated rapid deployment framework for RLW process is proposed to improve right-first-time implementation of RLW in assembly systems. The case study is an automotive door assembly.
SOLUTION - A framework is developed enabling closed-loop optimization of system layout, task assignment, fixture layout, process parameters, robot path planning and programming as an interlinked iterative approach. The proposed ‘Push’–‘Pull’ framework is necessary for rapid deployment. The individual key enabling technologies (KETs) have been implemented as a set of software modules within the RLW Navigator system. The software modules iteratively exchange their optimization results, design solutions and related KPIs, thus progressively updating the overall solution while at the same time keeping the coherence of the results provided by each software module. The control of the flow of information between these sub-modules and the optimization is performed by a workflow implemented within the commercial optimization software modeFRONTIER 4.5.
BENEFITS - System design revealed that the RLW joining process has several benefits over RSW and self-pierce riveting (SPR) joining. The number of robots in the overall production system was reduced, specifically, 5 robots in the RLW cell instead of 14 robots in the RSWcell. The floor space required for the production was reduced approximately by 50%. Equally important, the estimated total energy demand per product decreased by 57%. All engineering changes could be implemented by using the appropriate simulation platforms that resulted in 98% joints achieving right-first-time quality. This is in sharp contrast with the current industrial practice which can reach up to 55% success during the equivalent design stage