What does 2D conveyor tracking work with?

It works with FANUC iRVision, Cognex In Sight, Keyence vision, Conveyor encoders, FANUC R 30iB, ABB IRC5, Yaskawa DX200, Universal Robots controllers, Delta robots.

What should I watch for with 2D conveyor tracking?

Encoder resolution and mounting rigidity directly affect pick precision. Belt slip introduces position error over long conveyor runs and needs periodic checking. Camera trigger timing must be referenced to the encoder, not to a fixed clock. Work window boundaries have to account for robot deceleration at the exit edge. Multi-robot handoff on the same belt requires explicit part-ownership arbitration. Belt speed changes during a pick require the tracking software to keep the offset live. Downstream buffering must not assume a fixed cycle when picks are variable.

Vision

2D conveyor tracking

Vision-guided technique that lets a robot pick moving parts off a conveyor by tracking position through an encoder-synced camera trigger.

Conveyor tracking lets a robot pick moving parts without stopping the belt. A 2D camera detects the part upstream, an encoder on the conveyor shaft measures belt travel, and the robot controller combines the two to keep the part's position updated as a moving target frame.

The core piece is a robot-side software option (FANUC's Line Tracking, ABB's Conveyor Tracking, Yaskawa's MotoConveyor, and equivalent packages from KUKA and Universal Robots). These handle the coordinate transform between the fixed camera frame and the moving belt frame so the linear interpolation target updates in real time.

In a cell the flow is: camera triggers as the part crosses a photoeye or a defined encoder count, vision returns a part pose in the camera frame, the tracking software translates that into the belt frame, and the robot picks while continuously updating the target as the belt moves.

Reach for it in pick-and-place, packaging, and case-packing cells where stopping the conveyor kills throughput or where upstream product spacing is variable. It's common on food and consumer-goods lines where parts arrive in a random shuffle rather than a fixed pitch.

Selection depends on takt time, part size, belt speed, and how much of the belt the robot can actually reach at speed. Faster belts and smaller parts push toward high-speed delta robots, while heavier or slower work fits a standard six-axis arm like the FANUC LR Mate 200iD.

Encoder resolution, mounting rigidity, and belt slip drive the achievable pick precision, and they matter more than the vision camera specs in most failure modes. A slipping belt or a loose encoder coupler will show up as parts consistently missed on one side of the work window.

Commissioning pitfalls: camera trigger timing that isn't referenced to the encoder (so a belt speed change breaks the pick offset), work window boundaries that don't account for robot deceleration, and downstream buffering that assumes a fixed cycle when the tracking cell is actually variable.

Related integrations

fanuc irvision cognex in sight

Common questions

What does 2D conveyor tracking do?: Vision-guided technique that lets a robot pick moving parts off a conveyor by tracking position through an encoder-synced camera trigger.
What does 2D conveyor tracking work with?: It works with FANUC iRVision, Cognex In Sight, Keyence vision, Conveyor encoders, FANUC R 30iB, ABB IRC5, Yaskawa DX200, Universal Robots controllers, Delta robots.
What should I watch for with 2D conveyor tracking?: Encoder resolution and mounting rigidity directly affect pick precision. Belt slip introduces position error over long conveyor runs and needs periodic checking. Camera trigger timing must be referenced to the encoder, not to a fixed clock. Work window boundaries have to account for robot deceleration at the exit edge. Multi-robot handoff on the same belt requires explicit part-ownership arbitration. Belt speed changes during a pick require the tracking software to keep the offset live. Downstream buffering must not assume a fixed cycle when picks are variable.