Towards an integrated framework for modeling, control, and validation of production systems

Abstract: To prepare a production system before installation includes a multitude of computer assisted tasks. Virtual preparation, including many simulation supported activities, has made great strides during the last decade. Industrial robots are routinely programmed offline in simulation environments. Assembly sequences and material transportation are validated using simulation, as well as the ergonomics of human workers. In recent years, virtual commissioning has been added as another possible activity that can be performed using a simulated production system. The purpose of virtual commissioning is to enable the control software, which controls and coordinates the different devices in a production station, to be tested and integrated in a virtual environment before the system is installed. However, there is still a gap between developing the control system and the earlier simulation activities. This thesis presents steps taken towards bridging this gap, by integrating virtual commissioning into the existing virtual preparation workflow. This is done by formalizing how the control system development can be performed, and enabling testing of logic related to the coordination of devices to be performed during the entire virtual preparation work. Furthermore, it is shown how existing preparation work for industrial robots can be brought into this formalized workflow to enable reuse and a gradual transition from traditional preparation methods into the proposed workflow. This is done by automatically generating models of how existing robot programs behave, which allows the coordination of them to be optimized, verified, and simulated within the proposed workflow. Lastly, it is demonstrated how virtual reality technology can be used in combination with virtual commissioning to perform testing of safety protocols by ``physically'' interacting with safety equipment and simulated human-machine interfaces.