Operations, schedules & troubleshooting

Verify stable operation with one camera on the local network before adding cameras or connecting remote sites. Use the checks below for daily operation and troubleshooting.

Daily operator check

1. Check that the sidebar AI status is ready and important cameras have live video and fresh analyses.
2. Review saved critical incidents and snapshots since the previous shift, then check live analyses.
3. Check event delivery records in Settings → API & integrations and verify receipt at the destination.
4. Check that cameras have not been moved, blocked, out of focus, or blinded by backlight.

Remote camera connection failures

Remote RTSP connections can be blocked by the ISP, CGNAT, a firewall, or camera settings. Ask the network administrator to provide an on-site computer, site agent, or VPN so Watch can reach the camera over a private network. Keep RTSP off the public internet.

Schedules and reports

Schedules adjust monitoring missions during a selected time window. Keep Watch and camera monitoring running. The standard v0.39 interface does not offer scheduled text reports or cron jobs. Ask support about availability for your installation if you need another type of scheduling.

Event and snapshot retention

v0.39 live monitoring saves confirmed critical incidents and subsequent evidence. Safe and suspicious results remain in live analysis. Critical snapshots are retained for 14 days by default, with retention adjustable by an administrator. An event’s text may remain after its snapshot expires. Filter by camera and time, then check that an export contains the evidence you need before sharing it.

Accounts and camera permissions

Open Settings → Access control and check Account mode. Local mode does not require individual sign-in or enforce source rules. Ask your administrator or support to enable multi-user access if each person needs an account. User sign-in and the product license key are separate settings.

1. Once multi-user access is enabled, an administrator can enter an email, a password of at least eight characters, and a role under Create user, then select Create account. Viewers see granted content; operators can create and operate their own offline cases; admins manage users, source access, and integrations.
2. Expand Source access: sites, groups, and cameras. Configure sources and grant the appropriate scope to each user. Live, History/search, Export, and Manage are independent permissions; select each one the person needs.
3. When the rules are ready, select Enforce source permissions, press Save access, and check for Saved. Have the user sign in with their own account to verify camera visibility and export access.

For a missing camera or blocked export, check My effective access, then ask an administrator to review the account role, source grants, and case permissions. Saving rules in Local mode does not restrict other people’s access.

Update the desktop app

Open Settings → General → Safe application updates. Choose an update policy and release channel, then select Check for updates. Select Download when a version is available. Once downloaded, select Restart and install during an agreed break in monitoring. After Watch reopens, check the version, live video, and fresh analyses.

If an update fails, record the displayed error and Recovery state. A Restore button is available only when Watch has a version it can restore. If no update or restore option appears, follow the installation guide or ask support for the next step.

Troubleshoot by symptom

1. No video: check camera power, the local network, and RTSP URL, then see the RTSP guide.
2. Video but no analysis. Confirm Monitoring is enabled in the camera’s settings and the sidebar AI status is ready. For a connection check, open Settings → Camera AI engine, expand Analysis frequency and AI model connection, and select Test Local.
3. No notification. Check for a saved event, then inspect the connector’s source and severity filters, dry-run state, and delivery records.
4. Too many false alarms: specify the area, behavior, and exceptions in the mission, then repeat the same scene test.