Machine monitoring troubleshooting

Learn how to troubleshoot common issues with machine monitoring.

Machine attribute not updating

Symptoms: Machine attribute shows "No Data" or stale timestamp

Root causes:

1. OPC UA or MQTT connection lost

   • Navigate to Shop Floor > Machines > Machine Data Sources to view the Connector Host status
   • Click the Test button on the connector to test the connection
   • Verify network connectivity, including firewall rules and port forwarding

2. Attribute not mapped correctly

   • Navigate to Machine > Configuration tab
   • Verify the attribute maps to the correct OPC UA node or MQTT topic
   • Click Get Latest Message (MQTT) or browse nodes (OPC UA) to verify the data source publishes

3. Rate limit exceeded

   • Verify the total Hz across your instance does not exceed the 500 Hz limit
   • Reduce polling frequency or remove unnecessary attributes

Solution checklist

• Connector Host shows "Connected" status
• Machine attribute maps to an active data source
• Data source publishes data (verify in Kepware or broker UI)
• Machine is not in "Offline" state (verify in Machine Status widget)

Machine trigger not firing

Symptoms: Expected state change or activity field update does not occur

Root causes:

1. Trigger condition never met

   • Example: Trigger specifies "When Machine Current > 10" but the attribute maximum is 8
   • Navigate to Machine > Activity tab to verify attribute values

2. Trigger disabled

   • Navigate to Machine Type > Triggers and verify the toggle is ON

3. Multiple conflicting triggers

   • Example: Trigger A sets the state to Running, and Trigger B immediately sets the state to Stopped
   • Review all triggers on the machine type for conflicts

4. Trigger action references wrong machine

   • If using "at this station" in app triggers, verify the machine is assigned to the station

Solution checklist

• Trigger is enabled (toggle ON)
• Trigger condition is achievable (verify the attribute range)
• No conflicting triggers exist
• Station assignment is correct (if using "at this station")

OEE calculation incorrect

Symptoms: OEE percentage does not match expected values

Root causes:

1. Incorrect state mapping

   • The "Running" state must be configured for the Availability calculation
   • Navigate to Machine Type > States and verify "Running" is marked as "uptime"

2. Part count not incrementing

   • Verify the machine trigger writes to the "Part Count" activity field
   • View the Machine Activity table for part count values

3. Ideal Run Rate not set

   • The Performance calculation requires an ideal run rate
   • Set the ideal run rate using the "Ideal Run Rate" machine activity field or an app input

4. Clock skew

   • Machine timestamps are out of sync with the Tulip server
   • Verify timezone settings on the Connector Host and Tulip instance

Solution checklist

• "Running" state is configured as uptime in Machine Type
• Part Count activity field populates correctly
• Ideal Run Rate is set (either static or dynamic)
• Timezone alignment is verified

App not displaying machine data

Symptoms: Machine Status widget shows "No Machine Selected" or appears blank

Root causes:

1. Machine variable not populated

   • If using a variable datasource, verify the variable is set at runtime
   • Add a text widget to display the variable value for debugging

2. Station assignment missing

   • If using "at this station" logic, verify:
     • A station is selected in Player
     • The machine is assigned to that station

3. Machine permissions

   • The user role may not have permission to view machine data
   • Navigate to Settings > Roles and verify the role has "Machine Monitoring" permissions

Solution checklist

• Machine variable is populated (verify the variable value)
• Station is selected in Player (if using "at this station")
• Machine is assigned to the station (navigate to Shop Floor > Machines > Configuration)
• User has Machine Monitoring permissions (navigate to Settings > Roles)

Automation tasks exceeding budget or rate limit

Symptoms: Unexpected automation costs due to high task execution volume

Root causes:

1. Trigger fires too frequently

   • Example: An automation on "Every time machine attribute changes"
   • If the attribute updates at 1 Hz, that results in 3600 tasks/hour per machine

2. Missing conditional filters

   • The automation runs even when not needed
   • Add a condition: "Only when attribute value > threshold"

3. Redundant automations

   • Multiple automations perform similar tasks

Solutions

1. Batch operations

Before:
Automation: When machine attribute changes → Write to table
(Executes 3600 times/hour)

After:
Machine Trigger: When attribute changes → Update activity field
Automation: Every 1 hour → Copy activity field to table
(Executes 1 time/hour)

2. Add conditional filters

Before:
Automation: When machine state changes → Send notification
(Executes on every state change, even Running → Running)

After:
Automation: When machine state changes AND new state = "Stopped" → Send notification
(Executes only on Stopped state)

3. Use machine triggers for eligible tasks

• If the task only updates machine state or activity fields, use a machine trigger (free)
• Reserve automations for table writes, connector calls, and notifications

Further reading

• Machine Monitoring
• How to Add and Configure Machines
• Machine Monitoring Architecture
• Build Your First MQTT Data Source