Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
ISO/IEC 29341-24-2 defines the Temperature Sensor service, a fundamental component of the UPnP HVAC device architecture. This service provides standardized temperature measurement capabilities that feed data to the thermostat service (ISO/IEC 29341-24-10) and other HVAC control logic. The temperature sensor service is mandatory for all UPnP HVAC devices, as temperature feedback is essential for closed-loop climate control. The standard specifies accuracy requirements, update intervals, calibration procedures, and event notification thresholds to ensure reliable and consistent temperature data across the automation network.
The temperature sensor service defines strict accuracy requirements based on the application class. Class 1 sensors, intended for comfort control applications, must achieve an accuracy of ±0.5°C across the operating range of 0°C to 50°C. Class 2 sensors, suitable for general monitoring, require ±1.0°C accuracy. Class 3 sensors for coarse monitoring need ±2.0°C accuracy. The service reports both the temperature value and the sensor class, allowing control points to assess the reliability of the data and adjust their control algorithms accordingly.
| Sensor Class | Accuracy | Operating Range | Typical Application |
|---|---|---|---|
| Class 1 | ±0.5°C | 0°C to 50°C | Comfort HVAC control, thermostat feedback |
| Class 2 | ±1.0°C | -10°C to 60°C | General monitoring, energy management |
| Class 3 | ±2.0°C | -20°C to 85°C | Coarse monitoring, fault detection |
The service also defines the update interval, which determines how frequently the sensor refreshes its reading. Class 1 sensors must update at least every 10 seconds, Class 2 every 30 seconds, and Class 3 every 60 seconds. These update intervals are separate from the event notification threshold, which controls how frequently state change events are sent to subscribed control points. This separation allows sensors to sample frequently for internal use while limiting network traffic.
ISO/IEC 29341-24-2 supports multi-sensor configurations where multiple temperature sensor instances provide readings from different locations. The service does not define aggregation logic itself but provides mechanisms for control points to retrieve readings from individual sensors and perform their own aggregation. Common aggregation strategies include averaging all sensor readings (suitable for open-plan spaces), using the maximum reading (for cooling-dominated control), or using the minimum reading (for heating-dominated control).
Each temperature sensor instance exposes a unique sensor identifier, location description, and installation date in addition to the current temperature reading. This metadata allows control points to present meaningful information to users, such as displaying temperature readings with room labels. The service also supports a calibration offset parameter, which can be adjusted to compensate for known sensor biases or installation-specific factors such as proximity to heat sources.
The calibration offset feature is particularly valuable for retrofit installations where sensors cannot be ideally positioned. By applying a calibration offset, installers can compensate for known temperature biases. For example, a sensor mounted near a window that reads 2°C cooler than the room average can be adjusted with a +2.0°C offset. The calibration offset can be set programmatically through the service’s SetCalibrationOffset action and should be stored in non-volatile memory to survive power cycles.
An important engineering consideration is sensor fault detection. The service defines a sensor status state variable that indicates whether the sensor reading is valid, invalid, or the sensor has failed completely. Control points should monitor this status and switch to fail-safe temperature values or alternative sensor inputs when a sensor fault is detected. The recommended fail-safe behavior is to use the last valid reading for a maximum of 30 minutes, after which the system should enter a protective mode if no valid sensor data is available.