Overview

This topic explains how to set up, teach, validate, and run jobs on OmnnEye Autoview. It covers Monitoring, Configuration, and Statistics, with a quick checklist for first-run and a troubleshooting section. A downloadable manual is linked at the end.

Before you begin

• Confirm interlocks, stack light, and E stop.

• Load the correct recipe and apply saved positions.

• Clean lenses; verify stable lighting and camera exposure.

• Keep at least 10 good and 10 typical defect parts ready for validation.

Quick start (summary)

1. Create or Duplicate a program. Name it with part code and FOV.

2. Set camera exposure, trigger mode, and frame rate.

3. Capture a clean reference image.

4. Train Find Object on stable, unique features.

5. Add tools: measurement, code read (1D, 2D, GS1), OCR, defect checks.

6. Set tolerances: Nominal, minus Tol, plus Tol.

7. Validate on 30–50 mixed samples; tune limits to meet yield and takt.

8. Save and Export the program. Version on every significant change.

Operations

• Monitoring: Run, Stop, Load, Trigger; review image overlays and result rows.

• Statistics: Track yield, top defects, mean, sigma, and 3 sigma limits. Export CSV for audits.

• Data: Results are buffered for speed. Export regularly to persist values and images.

Troubleshooting

• No or dropped frames: lower frame/trigger rate; check exposure and CPU.

• Unstable measures: improve lighting, reduce glare, refine ROIs, recalibrate.

• Wrong orientation: retrain Find Object with better features and masks.

• Export fails: verify path, credentials, and network reachability.