In-Space Assembly and Manufacturing Using Synchronized Robotics

Articulated robots are being entrusted with an ever-increasing range of manufacturing operations across various sectors. This is stemmed, among other factors, by the broad movement range, flexibility, and small footprint of these robots. However, when it comes to high-payload operations, gantry robots, with their rigid construct, are still the option of choice. This is particularly true when precision is of concern. Unfortunately, a rigid gantry setup is difficult to implement in space especially considering weight constraints and the need for flexibility. Modular robotic kinematic systems offer process flexibility and diversity. Robotic kinematic systems are already being used for in-space applications. This proposal is aimed at developing a machine learning-based software solution that would allow for high-precision, high-payload manufacturing operations to be carried out autonomously using multiple articulated robots. The work targets robotic sheet metal forming, which given its versatility and simplicity is an ideal operation for both on-ground and in-space manufacturing/repair of metal parts. The robotic forming cell consists of two heterogenous, 6-axis robots mounted on two linear tracks and a real-time monitoring and control system. The robots must work in coordination with each other to incrementally form a sheet of metal into a part based on an input CAD file. This work introduces a software solution which addresses a number of challenges. First, both robots need to be synced while experiencing differing loading conditions imposed by the forming operation (e.g., due to variation in friction). Second, the positioning of the two robots need to be dynamically adjusted to make sure the overall compliance of the robotic system is minimized. Third, the system monitors the forming operation in real-time and assesses its deviation against the initially defined path. The feedback from the monitoring system needs to be actively used to update the predefined path.