Skip to content

FAQ

Does Visual Detector need a cloud connection?

No. Detection runs on the ctrlX device. Pictures, models and results stay in your network.

Which cameras work?

GenICam / GigE Vision industrial cameras, USB / UVC cameras, and RTSP or HTTP streams. Each connected camera needs one Camera Connector licence. See Camera Connector.

How does the PLC read results?

From the ctrlX Data Layer, in three steps: wait for ready, write trigger, read when picture/trigger shows your value. See Data Layer.

Do I need a model?

No. Color Detection finds parts by colour with no model. A model from InSiSo Model Bench is needed for Object Detection.

Can it give pick positions to a robot?

Yes, with the Machine Coordinate Transformation licence. See Coordinate Mapping.

What happens after a power cut?

The app starts by itself, reconnects its cameras, loads its model and detects, with no button pressed. The picture counter carries on where it stopped.

Why does the model run on the CPU when the device has a GPU or a Hailo module?

Diagnostics ▸ Accelerators says why, usually a driver version the snap does not carry. See Accelerators.

Why does the app show a login page?

ctrlX OS did not pass your sign-in to the app; the box above the login says why. Open the page from the ctrlX OS sidebar, not from a bookmark or a typed address. If the login still appears, log out of ctrlX OS, log in again and open the page from the sidebar. See Sign in.

Why does ready stay FALSE?

message says what to fix, for example a camera that is not connected or a licence that is not held. Diagnostics shows the same in its chips.

What should I send to support?

The file from Diagnostics ▸ Download diagnostics (.zip), the ctrlX device and ctrlX OS version, and what you did. See Support.