Example FANUC program files for the generic vision interface

Version: 1.0.0

This repository demonstrates how to use the Generic Vision Interface with wenglor Machine Vision Devices on a FANUC controller. The included KAREL library (w_library.pc) and TP programs form a working sample program that you can adopt and customize for your application.

NOTE

This repository focuses exclusively on FANUC Robots-specific topics. For general robot vision information, please refer to the wenglor robot vision manual.

📖 Full documentation is available in the online manual

---

Table of Contents

• Prerequisites
• Installation
• Running the Sample Program
• Configuration (W_LIBRARY KAREL variables)
  • Network & Socket Setup
  • Adjusting Parameters
  • Teaching Poses
• Troubleshooting
  • Communication Errors
  • Insufficient Calibration Accuracy
• Support & Feedback

---

Prerequisites

Tested with a FANUC R-30iB Mate Plus controller (software V9.40) and an LR Mate 200iD robot arm.

• Basic knowledge of TP and KAREL programming.
• A FANUC controller with the system variable $KAREL_ENB set to 1.
• Socket Messaging (Host Comm client) configured for the Machine Vision Device.
• A B60 or Machine Vision Controller (MVC).
• A uniVision job for calibration and object detection.

---

Installation

1. Get the files from the sources directory.
2. Transfer them to the robot controller via FTP or a USB stick (target device Mem Device (MD)).
3. Make sure $KAREL_ENB is set to 1.
4. Run W_LIBRARY once to initialize the KAREL variables (it returns a cam error on the first run — that is expected).
5. Follow the configuration steps.

---

Running the Sample Program

1. Select W_SINGLE_DETECT, W_MULTI_DETECT or W_UPDATE_REFERENCE_FRAME.
2. If no calibration data is present on the device, the calibration process starts automatically. Answer the prompts under Menu → User using Deadman + SHIFT + F1/F2.
3. Execute the program and monitor the messages on the teach pendant.

---

Configuration (W_LIBRARY KAREL variables)

Adjust the variables under Data → Karel Vars after selecting the W_LIBRARY program.

Network & Socket Setup

• Set the robot controller IP under Menu → (6) Setup → Setup 2 → (9) Host Comm → TCP/IP (e.g. 192.168.100.11).
• Configure the Host Comm client (e.g. C1) with the device IP (default 192.168.100.1) and port (default 32006).
• Make sure the client tag matches the w_client_tag KAREL variable (default C1:).

Adjusting Parameters

Click to see the relevant parameter adjustments in the w_library.pc file

Variable	Default	Note
w_client_tag	C1:	Host Comm client tag.
w_use_case	camera_on_robot	camera_on_robot or camera_not_on_robot.
w_calib_target	zvzj002	Calibration target (zvzj001–zvzj004).
w_calib_job	calibration.u3p	uniVision calibration job.
w_detect_pose_reg	65	Detection pose register (PR). Keep in sync with the TP programs.

See the User Configuration page for the full list.

Teaching Poses

Set up to eleven calibration poses (minimum five) in the W_CALIB_POSES array under Data → Karel Pos. Set the detection pose in PR[65] (or another PR, if you also update w_detect_pose_reg and the TP programs).

---

Troubleshooting

Communication Errors

• Verify IP/port in KAREL variables
• Ensure the robot server on the Machine Vision Device is active
  • Go to the device website → Jobs → Processing Instance → Robot Server
• Check network connectivity/firewall

Insufficient Calibration Accuracy

You can use more than 5 calibration poses by adding more calibration poses in the W_CALIB_POSES array under Data → Karel Pos.

Click to see where to set the poses in the w_library.pc file

Set up to eleven calibration poses (minimum five) in the W_CALIB_POSES array under Data → Karel Pos. For better accuracy, use seven to eleven poses with increased variation, especially in pose angles.

See the Troubleshooting page for more.

---

Support & Feedback

• Bugs: Please open a new Issue in the GitHub Issues section if needed.
• Feature Requests & Ideas: Discuss suggestions in the Discussions → Ideas category under GitHub Discussions.