What does Keyence vision work with?

It works with EtherNet/IP, PROFINET, Socket messaging, FANUC R 30iB, ABB IRC5, Universal Robots controllers, KUKA KR C4.

What should I watch for with Keyence vision?

Menu-driven tools hit their ceiling on novel or highly custom defect logic. Multi-camera stations need careful trigger synchronization to avoid corrupted inspections. Service and support routes through Keyence directly rather than open integrators. Recipes are portable between compatible controllers but may not carry across generations. Fieldbus tag mapping is configured separately from the vision recipe and needs its own documentation. Sensor and lighting selection is largely bundled with the controller purchase. Custom preprocessing outside the tool set generally can't be added later.

Vision

Keyence vision

Keyence CV-X and XG series vision systems used for high-speed inspection, measurement, and robot guidance with menu-driven setup.

Keyence's vision controllers (the CV-X and XG-X series) pair with the company's own cameras and lighting, and are set up through a menu-driven interface rather than a scripting or graph-based programming environment. That trades some flexibility for a much shorter learning curve for line technicians who don't come from a vision background.

In a robot cell they act as an external inspection or guidance station, similar in role to Cognex In-Sight. Results pass to the robot as coordinates or pass/fail flags over EtherNet/IP, PROFINET, or discrete I/O, depending on how the controller is optioned.

Multi-camera setups are common for 360-degree inspection or wrap-around dimensional checks. All cameras run off the same controller, which keeps their triggers coordinated and gives one place to configure the recipe.

Pick Keyence when the plant already runs Keyence sensors, when the vision recipe has to be maintainable by technicians rather than a dedicated vision programmer, or when Keyence's direct support model is worth more to the customer than the open toolchain a Cognex or Basler cameras build would give.

The trade-off is that menu-driven tools hit limits sooner than open programming. Novel defect logic, custom preprocessing pipelines, or unusual color-space work often can't be built inside the standard tool set, and there's no equivalent to writing a VisionPro script.

Commissioning issues cluster around trigger synchronization on multi-camera stations (one camera firing early or late will corrupt an inspection), lighting consistency across shifts, and the controller's tag map when it's connected to a PLC through a fieldbus like PROFINET or EtherNet/IP.

Typical uses: end-of-line quality inspection on a stamped or machined part, robot pick-and-place guidance where parts arrive on a tray, and inline assembly verification tied to a reject conveyor.

Related integrations

cognex in sight 2d conveyor tracking

Common questions

What does Keyence vision do?: Keyence CV-X and XG series vision systems used for high-speed inspection, measurement, and robot guidance with menu-driven setup.
What does Keyence vision work with?: It works with EtherNet/IP, PROFINET, Socket messaging, FANUC R 30iB, ABB IRC5, Universal Robots controllers, KUKA KR C4.
What should I watch for with Keyence vision?: Menu-driven tools hit their ceiling on novel or highly custom defect logic. Multi-camera stations need careful trigger synchronization to avoid corrupted inspections. Service and support routes through Keyence directly rather than open integrators. Recipes are portable between compatible controllers but may not carry across generations. Fieldbus tag mapping is configured separately from the vision recipe and needs its own documentation. Sensor and lighting selection is largely bundled with the controller purchase. Custom preprocessing outside the tool set generally can't be added later.