Trends, Controllers, Alarms and Fault Classification
An alarm is a signal to investigate, not a diagnosis. Advanced verification compares commanded state, measured response, independent evidence, spatial pattern and time-series behavior before classifying a condition as sensor, control, equipment, distribution or process-related.
Learning objectives
- LO-LH-TECH2-002-01
- LO-LH-TECH2-002-02
- LO-LH-TECH2-002-06
Key vocabulary
The controller or automation state the system is requesting, such as on, off, open, closed or a target output.
The measured physical response that should follow a command if the sensor, controller and equipment chain is working as expected.
A bounded evidence-based category describing where a problem is most likely located without claiming an unverified root cause.
Separate command from response
A controller display can show a command without proving the equipment produced the intended physical response. Compare independent process measurements, equipment indicators and trends.
Use trend shape and spatial pattern
Step changes, flatlined values, repeated cycling, delayed response and location-specific deviations can suggest different verification paths. The pattern narrows hypotheses but does not replace direct checks.
Lighting measurements need repeatable geometry
PPFD comparisons should use consistent canopy plane, sensor orientation and sampling locations. A single convenient point should not stand in for a spatial lighting map.
Classify without overstating root cause
Technician II may document a likely sensor/data-quality fault, control-response mismatch or equipment-response failure while escalating repairs or engineering decisions that exceed assigned authority.
Worked examples
- A dehumidifier command changes but room and local humidity trends do not respond; the condition is classified as a response mismatch pending authorized equipment verification.
- A PPFD complaint is tested with a repeatable map rather than one handheld reading near the aisle.
Common mistakes
- Assuming commanded state proves physical response.
- Resetting repeated alarms until they disappear.
- Using one light reading to characterize the whole canopy.
Practical application
Annotate a trend showing commands, alarms, independent measurements and relevant work events. Assign a provisional fault class and list the next safe check that would distinguish sensor, controller, equipment or spatial-process causes.
Lesson summary
Good fault classification connects commands to real response, uses spatial and temporal evidence, and stops short of claiming a repair diagnosis without adequate verification.