A novel multi-brand robotic software interface for industrial additive manufacturing cells

This study aims to report the development of a provisional robotic cell for additive manufacturing (AM) of metallic parts. To this end, the paper discusses cross-disciplinary concepts related to the development of the robotic cell and the associated command and control system such as the Computer-Aided Design (CAD) interface, the slicing software and the path planning for the robot manipulator toward printing the selected workpiece. This study also reports the development of a virtual production cell that simulates the AM toolpath generated for the desired workpiece, the adaptation of the simulation environments to enable AM and the development of a user application to setup, command and control the AM processes. If a digital twin setup is efficiently built, with a good correlation between the simulation environment and the real systems, developers may explore this functionality to significantly reduce the development cycle, which can be very long in AM applications where metallurgic properties, part distortion and other properties need to be monitored and controlled.

To generate the robot manipulator path, several simulation programs were considered, resulting in different solutions to program and control the robot of choice [in this study, Kuka and Asea Brown Boveri (ABB) robots were considered]. By integrating the solutions from Slic3r, Inventor, Kuka.Sim, Kuka.Officelite, RobotStudio and Visual Studio software packages, this study aims to develop a functional simulation system capable of producing a given workpiece. For this purpose, a graphical user interface (GUI) was designed to provide the user with a higher level of control over the entire process toward simplifying the programming and implementation events.

The presented solutions are compatible with the simulation environments of specific robot manufacturers, namely, ABB and Kuka, meaning that the authors aim to align the developments with most of the currently realized AM processing cells. In the long-term, the authors aim to build an AM system that implements a produce-from-CAD strategy i.e. that can be commanded directly from the CAD package used to design the part the authors are interested in.

This study attempts to shed light on the industrial AM, a field that is being constantly evolved. Arguably, one of the most important aspects of an AM system is path planning for the AM operation, which must be independent of the robotic system used. This study depicts a generic implementation that can be used with several robot control systems. The paper demonstrates the principle with ABB and Kuka robots, exploiting in detail simulation environments that can be used to create digital twins of the real AM systems. This is very important in actual industrial setups, as a good correlation between the digital twins (simulation environment and real system) will enable developers to explore the AM system in not only a more efficient manner, greatly reducing the development cycle but also as a way to fully develop new solutions without stopping the real setup. In this research, a systematic review of robot systems through simulation environments was presented, aiming to emulate the logic that is, used in the production cell development, disregarding the system brand. The adopted digital twin strategy enables the authors to fully simulate, both operationally and functionality, the real AM system. For this purpose, different solutions were explored using robots from two different manufacturers and related simulation environments, illustrating a generic solution that is not bound to a certain brand.

Using specific programming tools, fully functional virtual production cells were conceived that can receive the instructions for the movements of the robot, using a transmission control protocol/internet protocol. Conversion of the CAD information into the robot path instructions for the robot was the main research question in this study. With the different simulation systems, a program that translates the CAD data into an acceptable format brings the robot closer to the automatic path planning based on CAD data. Both ABB and Kuka systems can access the CAD data, converting it to the correct robot instructions that are executed. Eventually, a functional and intuitive GUI application capable of commanding the simulation for the execution of the AM was implemented. The user can set the desired object and run a completely automatic AM process through the designated GUI. Comparing ABB simulation with the Kuka system, an important distinction can be found, namely, in the exportation of the programs. As the Kuka program runs with add-ons, the solution will not be exported while maintaining its functionality, whereas the ABB program can be integrated with a real controller because it is completely integrated with modules of the virtual controller.

To conclude, with the solutions exploited, this study reports a step forward into the development of a fully functional generic AM cell. The final objective is to implement an AM system that is, independent of any robot manufacturer brand and uses a produce-from-CAD strategy (c.f. digital manufacturing). In other words, the authors presented a system that is fully automatic, can be explored from a CAD package and, consequently, can be used by any CAD designer, without specific knowledge of robotics, materials and AM systems.