Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
ISO/IEC IEC 29341-6-15 — UPnP Low Power — Part 6-15: Power Alert Service
Threshold-Based Power Event Notification for Intelligent Energy Management
The ISO/IEC 29341-6-15 standard defines the Power Alert Service for UPnP Low Power networks. This service provides a framework for devices to generate asynchronous alert notifications when power-related conditions cross user-defined thresholds. Unlike polling-based monitoring, the Power Alert Service uses event-driven architecture to immediately notify control points of critical power events, making it essential for demand-response automation, fault detection, and safety applications.
Use the Power Alert Service in combination with Energy Monitoring (6-14) to create a closed-loop energy management system: monitoring detects the condition, alerts notify the control point, and Power Control (6-17) executes the corrective action.
Alert Architecture and Threshold Configuration
The service defines a flexible threshold system where each alert condition is configured with upper and lower bounds, hysteresis, and enable/disable control. The key state variables are:
| State Variable | Type | Description |
|---|---|---|
| AlertConfiguration | string (XML) | XML-encoded list of alert rules with threshold values, hysteresis, and enable flags |
| AlertStatus | string | Current alert status: “normal”, “highAlert”, “lowAlert”, “fault” |
| AlertDescription | string | Human-readable description of the most recent alert condition |
| AlertTimestamp | string (ISO 8601) | Timestamp of the most recent alert trigger |
| AlertCount | ui4 | Number of alert events since last reset |
Event-driven alerting reduces network bandwidth consumption by up to 90% compared to polling-based monitoring, since messages are only sent when thresholds are crossed rather than at fixed intervals.
Alert Actions and Event Notification
The service provides the following actions for managing alert configuration and retrieval:
- GetAlertConfiguration() — Returns the current XML alert rule set. Control points parse this to render threshold configuration UI.
- SetAlertConfiguration(NewAlertConfiguration) — Replaces the alert rule set with a new XML configuration. The device validates the XML before accepting it.
- GetAlertStatus() — Returns the current aggregate alert status, description, and timestamp in a single call.
- ResetAlertCount() — Resets the AlertCount variable to zero, typically after an operator has reviewed and acknowledged the alerts.
The true power of the Power Alert Service lies in its eventing mechanism. When an alert threshold is crossed, the device multicasts a PowerAlert event to all subscribed control points. The event payload includes the variable that triggered the alert, the actual value, the threshold that was crossed, and a severity level (info, warning, critical). This allows control points to prioritize response actions.
When configuring multiple alert thresholds on a single device, ensure hysteresis bands do not overlap. Overlapping hysteresis can cause rapid toggling (thrashing) between alert states, generating event storms that degrade network performance.
Practical Deployment Scenarios
Demand-Response Integration. A utility company’s demand-response server can use the Power Alert Service to listen for “high demand” alerts from a whole-home energy monitor. When the alert triggers (e.g., power draw exceeds 10 kW for 5 minutes), the server dispatches a load-shedding command through the Power Control Service. This provides sub-minute response times without the overhead of cloud polling.
In safety-critical applications (e.g., EV charger over-temperature or arc-fault detection), the Power Alert Service must not be the sole alert path. Always implement a hardware-level interrupt or watchdog that can disconnect power independently of the UPnP stack, which may fail during a fault condition.
Fault Detection and Diagnostics. Industrial equipment can use the Power Alert Service to signal abnormal power consumption patterns. For example, a gradual increase in steady-state current beyond a configurable threshold may indicate bearing wear or refrigerant loss. The alert timestamp and description fields provide diagnostic context for predictive maintenance systems.
Frequently Asked Questions
Q: How many alert thresholds can a single service instance support?
A: The standard does not mandate a maximum. Practical implementations typically support 8–16 thresholds. The limiting factor is the XML configuration size, which should stay under 4 KB to fit within a single UPnP action response.
A: The standard does not mandate a maximum. Practical implementations typically support 8–16 thresholds. The limiting factor is the XML configuration size, which should stay under 4 KB to fit within a single UPnP action response.
Q: Can alerts be persisted across device reboots?
A: Yes. AlertConfiguration and AlertCount should be stored in non-volatile memory. AlertStatus returns to “normal” after reboot since the actual power condition must be re-evaluated.
A: Yes. AlertConfiguration and AlertCount should be stored in non-volatile memory. AlertStatus returns to “normal” after reboot since the actual power condition must be re-evaluated.
Q: What is the minimum hysteresis required to prevent alert thrashing?
A: A hysteresis of at least 5% of the threshold value is recommended. For a 1000 W threshold, use a 50 W hysteresis band (alert triggers at 1000 W, clears at 950 W).
A: A hysteresis of at least 5% of the threshold value is recommended. For a 1000 W threshold, use a 50 W hysteresis band (alert triggers at 1000 W, clears at 950 W).
Q: Does the service support alert aggregation from multiple sensors?
A: Not directly. Each service instance corresponds to one metering point. To aggregate, create a virtual device that collects alerts from multiple child services and re-publishes them as a consolidated alert service.
A: Not directly. Each service instance corresponds to one metering point. To aggregate, create a virtual device that collects alerts from multiple child services and re-publishes them as a consolidated alert service.
