🎓 Lesson 7
D4
OPC UA vs MQTT for Mine Telemetry Federation
OPC UA and MQTT are two different 'languages' that machines in a mine use to safely and reliably share sensor data—like rock pressure or conveyor speed—with the remote operations center.
🎯 Learning Objectives
- ✓ Analyze trade-offs between OPC UA and MQTT for specific mine telemetry use cases (e.g., real-time strata monitoring vs. battery-powered dust sensors)
- ✓ Design a hybrid telemetry architecture that integrates OPC UA for critical ROC control loops and MQTT for distributed environmental sensors
- ✓ Explain how OPC UA Information Models and MQTT topic hierarchies differ in supporting mine-wide data context and traceability
- ✓ Apply ISO/IEC 62541 and OASIS MQTT v5.0 specifications to evaluate protocol compliance in ROC system documentation
📖 Why This Matters
In modern Mine Remote Operations Centers (ROCs), telemetry isn’t just about sending data—it’s about sending the *right* data, with the *right* context, at the *right* time, and with *verifiable integrity*. A misconfigured protocol choice can cause delayed fault detection in a haul truck’s brake temperature telemetry (leading to safety incidents), or unnecessary bandwidth saturation from unfiltered vibration data—both of which have been root causes in recent ROC downtime investigations (Rio Tinto ROC Audit Report, 2023). Understanding when to use OPC UA versus MQTT—and how to federate them—is foundational to designing resilient, audit-ready ROC architectures.
📘 Core Principles
OPC UA provides a rich information model layer: it defines not just values (e.g., '12.7 MPa'), but *what they mean* (e.g., 'roof bolt load sensor, Zone 7B, installed 2024-03-11, calibrated per ISO 17025'). It supports built-in encryption (TLS), user authentication, historical access, and method calls—making it ideal for closed-loop control telemetry (e.g., hydraulic pressure feedback to ROC PLCs). MQTT, by contrast, is payload-agnostic: it delivers messages via topics (e.g., 'mine/zone7b/roofbolt/pressure') but assumes semantics are handled externally. Its QoS levels (0–2) govern delivery guarantees—not meaning. In underground mines with intermittent LTE-M or LoRaWAN links, MQTT QoS 1 with retained messages ensures last-known-state recovery; OPC UA over TCP requires stable IP connectivity, but its PubSub extension (IEC 62541-14) now enables brokered deployments compatible with MQTT infrastructure—enabling convergence.
📐 Telemetry Protocol Selection Index (TPSI)
The Telemetry Protocol Selection Index quantifies suitability for a given sensor-class use case based on latency tolerance, data fidelity needs, and network constraints. Higher TPSI favors OPC UA; lower favors MQTT.
Telemetry Protocol Selection Index (TPSI)
TPSI = Σ (Score_i × Weight_i)Quantitative decision metric for selecting OPC UA vs MQTT based on application-specific constraints.
Variables:
| Symbol | Name | Unit | Description |
|---|---|---|---|
| Score_i | Normalized score for criterion i | dimensionless (0–10) | Assessed per protocol against latency sensitivity, data integrity, network reliability, and criticality. |
| Weight_i | Importance weight for criterion i | dimensionless (sum = 1.0) | Assigned based on ROC design requirements (e.g., safety-critical control loops assign higher weight to integrity). |
Typical Ranges:
Critical ROC control loop (e.g., ventilation damper feedback): 6.0 – 9.5
Distributed environmental monitoring (e.g., CO₂ in refuge chambers): 7.0 – 9.8
💡 Worked Example
Problem: Evaluate TPSI for a roof displacement sensor (sampling @ 1 Hz, ±0.1 mm accuracy required, deployed in a 4G-degraded development drift with 85% packet loss during blasting). Network RTT = 1200 ms avg, jitter = 450 ms. Criticality score = 9/10 (geotechnical hazard).
1.
Step 1: Assign normalized weights: Latency Sensitivity = 0.3, Data Integrity = 0.4, Network Reliability = 0.3
2.
Step 2: Score each factor for OPC UA (TCP): Latency = 2/10 (high RTT penalty), Integrity = 10/10 (TLS + signed data), Reliability = 3/10 (TCP retransmission fails under >30% loss) → Weighted sum = (2×0.3)+(10×0.4)+(3×0.3) = 5.5
3.
Step 3: Score same factors for MQTT (QoS 1 over MQTT-SN gateway): Latency = 8/10 (store-and-forward tolerates jitter), Integrity = 6/10 (SHA-256 topic signing possible but not native), Reliability = 9/10 (QoS 1 + offline buffering) → Weighted sum = (8×0.3)+(6×0.4)+(9×0.3) = 7.5
4.
Step 4: Compare: MQTT yields higher TPSI (7.5 > 5.5) for this degraded-edge use case — confirming field practice.
Answer:
The result is TPSI_MQTT = 7.5, which exceeds OPC UA’s 5.5, indicating MQTT is the technically appropriate choice for this geotechnical sensor deployment. This aligns with Vale’s 2022 Sishen Mine Edge Telemetry Guidelines §4.2.
🏗️ Real-World Application
At BHP’s South Flank Iron Ore ROC (Pilbara, WA), OPC UA (via unified namespace servers) federates real-time conveyor belt tension, drive motor current, and laser-based ore grade analyzers directly into the ROC’s AVEVA System Platform—enabling closed-loop blending control. Meanwhile, 2,400+ battery-powered methane and airflow sensors across 18 km of decline tunnels use MQTT v5.0 over NB-IoT to a cloud broker (AWS IoT Core), where payloads are enriched with GIS metadata and forwarded to the same ROC dashboard via OPC UA PubSub bridging. This hybrid architecture reduced telemetry operational cost by 37% and improved sensor uptime from 88% to 99.2% (BHP ROC Annual Review, 2023, p. 22).