Researchers pair digital twin, cobot for remote AM finishing

Key Takeaways

• A cyber‑physical system integrates a Unity‑based digital twin, XR interface, a UR30 collaborative robot and a Mass Finishing RF‑50 centrifugal finisher to enable remote programming, supervision and intermittent tele‑operation of metal‑AM post‑processing.
• Operators can stay ≥10 m away from a noisy (≈85 dB A) and dusty finishing cell while retaining full control over robot joint trajectories and gripper actions.
• The digital twin validates every commanded pose against joint limits and collision models before MQTT messages reach the physical robot, adding a safety layer beyond the robot controller.
• Lab trials show the workflow is viable for small‑batch and one‑off production, cutting on‑site labor exposure and shortening set‑up time by up to 30 % compared with conventional manual finishing.

Remote Finishing: From Concept to Lab Prototype

Researchers from UNC‑Charlotte’s Digital Engineering for Advanced Manufacturing Laboratory and University of Oulu have built a fully‑connected cyber‑physical loop that lets a single operator program, monitor, and intermittently tele‑operate a finishing robot without stepping inside the cell. The work, published on September 8, 2023, targets the persistent bottleneck where highly automated metal additive manufacturing (AM) meets a manual, labor‑intensive finishing stage.

Core Hardware

Digital Twin Architecture

The twin lives as a WebGL application built in Unity, accessible from any browser or XR headset. It mirrors the robot’s kinematics, the finisher’s geometry, and the workpiece model. Operators can:

• Drag‑and‑drop waypoints, adjust joint angles, and open/close the gripper.
• Simulate the entire motion path, checking for collisions against a pre‑computed mesh of the finishing bowl.
• Record and replay trajectories, then push validated commands to the physical robot via MQTT.

Before transmission, the twin runs a constraint‑checking algorithm that enforces joint limits (±180° for axes 1‑3, ±120° for axes 4‑6) and a minimum 50 mm clearance from the bowl wall. The robot controller adds a second safety net, rejecting any command that violates its own internal limits.

Operator Experience & Safety Gains

Finishing with the RF‑50 generates high‑frequency vibration, metal dust, and sustained 85 dB A sound, conditions that can cause hearing loss and ergonomic strain after 4 h of exposure. By moving the operator to a remote workstation:

• Noise exposure drops to <40 dB (typical office level).
• Physical fatigue from standing in a vibrating environment is eliminated.
• Visibility improves dramatically; the XR view can render the part from any angle, overcoming the limited line‑of‑sight in the real cell.

Comparison: Conventional Manual Finishing vs. Remote Digital‑Twin Workflow

Enabling Small‑Batch and One‑Off Production

Because the digital twin can store, edit, and reuse process recipes, manufacturers can quickly switch between part geometries without re‑tooling the finisher. The system’s intermittent tele‑operation mode lets the operator intervene only when a deviation is detected, otherwise the robot runs autonomously. Early lab data indicate a 30 % reduction in overall lead time for batches of ≤20 parts compared with fully manual handling.

Bottom Line

The UNC‑Charlotte/University of Oulu cyber‑physical platform demonstrates that a digital twin‑driven, XR‑enabled cobot can safely relocate human oversight away from noisy, dusty AM finishing cells while preserving full control over the process. By validating robot poses in a virtual replica before execution and leveraging MQTT for low‑latency command delivery, the solution delivers measurable gains in safety, efficiency, and flexibility—particularly for low‑volume, high‑mix metal‑AM production. As manufacturers seek to close the gap between automated build and manual post‑process, remote twin‑assisted finishing offers a scalable pathway toward fully digitized, operator‑light factories.