Dynamic Vibration Compensation for Industrial and Collaborative Robot Arms

The problem

Rapid point-to-point motion can leave residual vibration at the TCP. Robot controllers compensate for rigid-body dynamics and configured payload properties, but they may not fully model joint and transmission compliance, link flexibility, and how these vary with configuration and payload. The resulting TCP oscillation increases settling time — the interval between the end of commanded motion and the TCP entering and remaining within its specified tolerance band.

Common mitigations include reducing speed, acceleration, or jerk; smoothing trajectories; and adding endpoint dwell. These measures can reduce throughput by lengthening motion or waiting for vibration to decay.

How Shaper works

Shaper adds a model-based feedforward layer ahead of the manufacturer's closed-loop servo controller:

1. Calibrate — The robot executes controlled motions while an accelerometer records its dynamic response. This characterizes the vibration modes of the individual arm — not a generic model for that platform.

2. Identify — Calibration data is processed through Reforge's cloud API to generate a configuration-dependent vibration model for that specific robot.

3. Control — The ShaperInterface sits between your trajectory planner and the robot SDK. It modifies outgoing joint position, velocity, and acceleration commands to reduce excitation of the identified flexible modes and suppress the resulting residual vibration.

The manufacturer's internal servo controller remains responsible for closed-loop joint control. Shaper is a feedforward layer — it does not require vibration feedback during normal operation.

The identified model is reusable across runs without recalibration. Material changes to payload, tooling, mounting, or mechanical condition should trigger revalidation. If performance falls outside acceptance criteria, the model can be fine-tuned or recalibrated.

Performance

• Greater than 80% vibration reduction across deployments (Reforge data)

• Up to 2x throughput improvement by eliminating settling time overhead, with net cycle time improvement accounting for any command-shaping delay

• Supports full-trajectory, residual-tail, fixed-window, and sample-by-sample processing modes

Integration

• Sits between your trajectory planner and the robot manufacturer's SDK — no firmware changes, no robot replacement

• Calibration requires temporary accelerometer mounting; Reforge provides a hardware kit

• Currently supported on Standard Bots, UFACTORY, Trossen, and Denso platforms

• Cloud API and browser-based interface available; Docker-based deployment supported for production environments

Business context

Settling time reduction translates directly to throughput when robot motion is the cycle-time constraint. In an application where Shaper eliminates two seconds of endpoint dwell per cycle, across 1,000 daily cycles and 260 operating days, it can free approximately 144 robot-hours of annual production capacity per arm. Shaper deploys without robot replacement, controller firmware changes, or permanent mechanical modification.