Troubleshooting

Before investigating, record the organization, project, device, point, and time. Hide passwords, security codes, and full access tokens in screenshots and logs.

A device cannot be added

Check its device ID, security code, and target organization. An already-bound device needs an Owner with assignment permissions to resolve ownership. Creating another project with the same name does not fix binding. For quota errors, check plan usage.

A device is offline or data has stopped updating

Check power, network connectivity, SIM status, antenna, and last connection time in that order. A device in the list is bound, but a past connection does not prove it is online now.

If it is online but a point has no value, check point enabled status, serial settings, slave ID, function code, address, and data index. See Modbus configuration.

Readings are incorrect

Symptom	Check
Value is 10 or 100 times too large or small	Scale factor and unit
Consistent offset	Offset value
Wrong negative or very large values	Signed format, byte order, index
Values appear on the wrong channels	Start address, register width, index

Compare readings with the sensor display or a reference instrument before changing settings. Do not widen thresholds to hide a decoding error.

Live values exist but history is empty

Check the range and project time zone, point creation time, debug mode, and historical storage. Live values and history use different storage paths, so live updates do not prove records were saved.

A GPS point has no track

Check that valid coordinates were reported, that the range contains positioning records, and that the map loads. Creating a GPS type does not generate location data.

Configuration or control is unavailable

Check role and project scope, plan status, and device support. Read-only access to an open project does not grant control.

A command was sent but equipment did not operate

Check the target point, command record, connectivity, and reported state, then inspect field wiring, the actuator, interlocks, and limits. After a timeout, check state before sending again. See Control devices for result meanings.

A schedule or rule did not run

• Check enabled status and whether the target point still exists and is compatible.
• For schedules, check project time, trigger mode, and active window.
• For rules, check the condition, continuing data reports, and rule health.
• Inspect execution records to distinguish failure, cooldown skips, and an already-satisfied target state.

See Schedules and Automation rules.

An alert remains active

Check whether new reports satisfy the recovery condition. Missing, stale, or offline data does not establish recovery. Acknowledge records human awareness only; see Alert states.

If a reading has just crossed a boundary without an alert, check subsequent reports. Both triggering and recovery require a qualifying condition across valid data for at least 30 seconds. A brief change does not immediately change the alert state.

In-app notifications arrive but email or SMS does not

Check notification settings: enabled status, destination, selected types, delivery results, and plan status. Verify tests and business messages separately. Restoring external delivery does not replay messages from a paused period.

An export fails or the file is empty

For a failure, inspect its error and reduce the range or point count. For a completed empty file, check the range and historical records. An expired file cannot be downloaded from its old link. See Data export.

Forwarded data does not arrive

Check enabled status and new device reports, then destination, credentials, and receiving logs. For MongoDB, also check the collection named after the organization slug. For InfluxDB, check the measurement and time range. See Data forwarding.

The API rejects access or cannot find a resource

For 401, check authentication and expiry. For 403, check role, scope, and plan. For 404, check the path, numeric key, and internal identifier. Start with a read-only request; see Use the Open API.