Future of Industrial Robot Control

robots are becoming commodity items for repetitive tasks in structured environments. For operations in cluttered or unstructured environments and sensor guided motion (using, e.g., proximity, vision, contact, and force sensors), motion planning is needed, using open-source packages or local optimization algorithms. The execution of the motion plan may be through either a streaming setpoint interface or a robot motion program consisting of a sequence of motion primitives. Streaming interface is useful in lower speed/precision applications such as visual servoing, compliance force control, and human-robot interaction. For higher performance industrial applications, both options have drawbacks. Streaming interface has low sampling rate, latency, and nonlinear dynamics. Robot motion program has to account for motion blending and robot-specific constraints. Future industrial robot control needs to address such challenges through a combination data analytics, robot modeling, and controller adaptation. We will present several different types of compensation schemes and demonstrate their applications in deep rolling and cold spray additive manufacturing.