What this page solves

Grid IoT nodes extend sensing and visibility into places where traditional meters and protection relays cannot easily reach. Small, rugged terminals sit on pole-tops, transformers and cabinets, watching conductor temperature, sag, ice load, cabinet conditions and simple status contacts. Each node must run for many years from a primary battery and harvested energy while sampling local sensors, running ultra-low-power firmware and pushing compact data packets over LPWAN or cellular links into SCADA, DMS or cloud analytics platforms.

• Extend sensing to pole-tops, transformers, switchgear, LV panels and street cabinets.
• Work under tight power budgets with primary batteries and energy-harvesting sources.
• Combine sensing AFEs, ultra-low-power MCUs and LPWAN or cellular modems in one node.
• Complement revenue meters and protection IEDs by adding more eyes and ears at the grid edge.

Typical deployment & node types

Grid IoT nodes appear wherever extra sensing is valuable but installing full meters or IEDs would be impractical. Different node archetypes share a common building block platform but face very different mechanical, environmental and power constraints. Pole-top line nodes live on towers and feeders; cabinet and switchgear nodes work inside metal enclosures; transformer and asset nodes monitor critical equipment; LV panel nodes sit in distribution or street cabinets watching downstream feeders and loads.

Pole-top line node. Mounted on overhead lines or towers, this node watches conductor temperature, sag and ice load, sometimes combined with vibration or tilt sensing. Power is usually scarce, so designs rely on primary batteries and compact solar or line-powered harvesting stages. Radio links favour LPWAN options such as LoRaWAN or NB-IoT to reach gateways or public networks from remote rights-of-way. Detailed sag and ice algorithms and alarm strategies are reserved for the Line Monitoring page.

Switchgear / RMU node. Installed inside metal-clad switchgear and ring main units, this node tracks internal cabinet temperature, humidity, partial discharge indicators and door or interlock status. Auxiliary DC supplies may be available but cannot be assumed under all fault and outage conditions, so low standby current remains essential. Cellular LPWAN such as NB-IoT or LTE-M is often preferred to penetrate cabinets and buildings. Cabinet-specific diagnostics are covered in switchgear and RMU monitoring topics.

Transformer / asset node. Attached to distribution transformers or other key assets, this node focuses on oil and winding temperature, tank vibration and acoustic noise, sometimes combined with bushing or terminal temperature. Power options range from auxiliary station supplies through thermal or vibration harvesting. Wireless choice depends on substation layout and reachability. Transformer-specific health indicators and analytics are developed further on the Transformer Monitor page.

LV panel / street cabinet node. Located in low-voltage distribution panels and street cabinets, this node monitors branch currents, neutral or earth leakage and hotspot temperatures together with basic environmental conditions. Access to AC power is usually easier but still subject to outages and wiring constraints, so the same low-power hardware platform is reused with adapted front-ends and modems. Detailed use cases around feeder imbalance, overload and leakage are addressed in dedicated LV panel monitoring pages.

Sensing front-ends (AFEs)

A grid IoT node aggregates very different sensor types into a small analogue and mixed-signal front-end. Temperature, strain and vibration channels sit next to electrical current and voltage pick-ups and digital environmental sensors. Each path has to survive lightning-induced surges, high common-mode voltages and long cables while still feeding clean, band-limited signals into high-resolution converters. The goal is not to implement application-specific algorithms in the node, but to provide robust, reusable AFE templates that can be mapped to line monitoring, transformer health or LV panel pages without redesigning the hardware every time.

Analogue sensor paths. NTC and RTD temperature probes, strain gauges and MEMS accelerometers form the analogue sensor group. These channels typically use bridge or divider networks, low-noise instrumentation amplifiers, differential gain stages and anti-alias RC filters in front of 16–24 bit converters. Layout must keep sensor leads away from noisy digital and RF nodes, and provide basic ESD and surge clamps at the enclosure boundary.

Electrical measurement paths. Small current transformers, shunts, voltage dividers and Rogowski coils feed AFEs that see high common-mode voltage and fast transients. Protection networks, creepage and clearance, burden resistor sizing and bandwidth limitation become critical. Front-ends must reject common-mode noise and include multi-stage surge protection so that high-energy events do not reach the ADC. Sampling bandwidth should be set high enough for the target phenomena but low enough to control aliasing and noise.

Digital sensor interfaces. Digital sensors on I²C, SPI or UART simplify local analogue design but bring bus-integrity and isolation questions. Pull-ups, bus extenders or isolated transceivers may be needed when sensors sit in noisy or floating domains. ESD protection, series resistors and common-mode chokes help keep fast edges under control. A mix of low-noise instrumentation amplifiers or op amps, 24-bit sigma-delta converters and analogue multiplexers can be re-used across line, transformer and LV cabinet monitoring use cases, while the application-specific thresholds and analytics stay on dedicated pages.

ULP MCU / SoC architecture

The controller at the heart of a grid IoT node spends almost all of its life asleep. It only wakes to sample sensors, perform light processing and push compact packets over a radio, then returns to deep sleep. A suitable MCU or SoC architecture therefore combines an efficient 32-bit core with multiple low-power modes, always-on domains for RTC and wakeup logic, integrated ADCs and timers, and secure peripherals, all arranged so that high-current activities remain short and infrequent.

Core and power domains. Cortex-M0+ and Cortex-M33 devices are common choices, with a main core domain and a tiny always-on island hosting the RTC, wake controller and watchdog. Deep-sleep modes disable clocks and memories that are not needed between measurements, targeting microamp average currents when combined with long duty cycles. ADCs, comparators and timers can be clocked only when necessary, and DMA engines can move samples without keeping the core awake.

Peripherals for sensing and connectivity. Integrated low-power ADCs handle modest channel counts, while external 24-bit converters connect through SPI for higher precision. Multiple SPI, UART and I²C ports link AFEs, LPWAN modems and digital sensors. GPIO leakage and pull configurations matter because many pins connect to long field cables. A security engine with AES, SHA and optional TrustZone-M or PUF-based key storage helps enforce secure boot and encrypted links without large energy overheads.

Firmware responsibilities. Firmware modules schedule sampling, compute basic features, check simple thresholds and pack measurements into efficient payloads. Link management code controls attach, transmit and sleep cycles on the LPWAN or cellular modem so that retries and error handling do not drain the battery. A small non-volatile storage layer on FRAM, EEPROM or Flash stores configuration, counters and diagnostic logs across outages. Over-the-air update mechanisms are designed as rare, well-planned events, with explicit checks against available energy and link quality. Long-term fleet analytics and health scoring remain in higher-level Telemetry and Asset Health systems.

LoRa / Cellular / NB-IoT – wireless link selection & interfaces

Grid IoT nodes typically rely on low-power wide area links instead of wired backhaul. Wireless choices determine not only coverage and operating cost but also power-tree sizing, connector options and PCB layout around high-voltage equipment. LoRa and LoRaWAN work well for private, gateway-based deployments; NB-IoT and LTE-M use operator networks for wide geographic spread with small telemetry payloads; legacy 2G, 3G or 4G modems remain in long-lived projects as fallback or migration paths and must still be considered when defining footprints and interfaces.

LoRa / LoRaWAN. LoRa and LoRaWAN suit clusters of nodes around substations, feeders or city districts where dedicated gateways can be installed. Hardware usually combines a low-power RF transceiver or SoC with an SPI or UART connection to the MCU. Link budgets are strong, but antenna placement and gateway siting dictate real performance. Private networks reduce dependency on operators but shift responsibility for gateway power, backhaul and maintenance to the utility or integrator.

NB-IoT / LTE-M. NB-IoT and LTE-M allow nodes to ride on operator infrastructure, fitting widely scattered assets and small, periodic telemetry. Modules draw hundreds of milliamps during attach and transmit bursts, so local power rails and bulk capacitors must be carefully dimensioned. UART is still the dominant host interface, with AT commands and a handful of control pins for power, reset and wake. Coverage, roaming and tariff structure become as important as RF performance when planning fleet deployments across large regions.

Legacy 2G/3G/4G and board-level hooks. Legacy cellular modems remain in the field and may provide a practical fallback where new radio profiles are not yet deployed. Designs often re-use the same UART, SIM and antenna footprint across LoRa and cellular variants so that future module changes do not require a new PCB. Nodes should expose signal-quality feedback such as RSSI, RSRP and SNR, include an eSIM or SIM holder area and implement RF test access for production and field diagnostics. All RF paths need careful ESD and surge protection, safety spacing and antenna siting away from high-voltage parts.

Energy-harvesting PMIC & power tree

Long-life grid IoT nodes cannot rely on primary batteries alone, especially when truck rolls and outage windows are constrained. Practical designs harvest small amounts of energy from sunlight, line current or temperature and vibration gradients, then buffer that energy in supercapacitors and secondary cells. An energy-harvesting PMIC sits between the sources and the storage, handling cold start, impedance matching or MPPT, overvoltage and undervoltage thresholds and safe charging profiles so that the node can survive seasonal and load-related variations.

Energy sources and PMIC requirements. Small pole-top solar panels provide daytime power with strong dependence on orientation, so the PMIC benefits from MPPT or simplified tracking. CT or Rogowski-based harvesters ride on line current and only produce power when feeders carry load, demanding careful coupling and isolation. Thermal and vibration harvesters around transformers and machinery deliver very low voltages and power levels, making cold-start performance critical. The same PMIC often has to manage these sources without disturbing measurement circuits or protection hardware.

Storage and power domains. Harvested energy is accumulated in a combination of supercapacitors and rechargeable cells. The supercapacitor delivers fast bursts for radio transmissions and startup events, while the cell provides long-term capacity. Downstream regulators derive separate rails for analogue AFEs, the MCU and the RF module, allowing the node to keep sensing and logging functions alive even when radio activity must be throttled. Protection functions in the PMIC enforce charge balancing, overvoltage and undervoltage limits and disconnect non-essential loads when storage becomes depleted.

Power budget and duty cycle. The power tree is dimensioned around a duty cycle of microamp sleep currents, millisecond-scale sensing windows and second-scale transmit events. Average harvested energy per day has to exceed or at least match this load to achieve multi-year life. When available power drops below budget, the node can reduce reporting frequency, disable non-critical sensors or postpone firmware updates. This page focuses on milliwatt-level harvesting and storage, while substation UPS topics cover kilowatt-scale backup power for protection and automation systems.

Mechanical, ruggedness & safety hooks

A grid IoT node must survive years of exposure on poles, cabinets and transformer surfaces while remaining safe around high-voltage equipment. Enclosure and connector choices define ingress protection, corrosion resistance and UV stability, while PCB-level measures such as creepage, conformal coating and controlled surge paths determine how well the electronics tolerate humidity, pollution and lightning-induced stress. This section focuses on node-level mechanics and ruggedness; system-level surge coordination and earthing are covered separately in the EMI / Surge / Lightning Protection page.

Enclosures, IP and environment. Outdoor nodes typically target IP65 or IP67 housings with UV-stable plastics or coated metals, stainless or corrosion-resistant fasteners and seals rated for repeated thermal cycling. Salt-mist and industrial pollution require careful choice of gasket materials and surface treatments, so that lids, cable glands and brackets do not corrode or crack long before electronics reach end of life. Internal layouts should avoid pockets where moisture condenses and should provide clear drainage and venting paths where appropriate.

Connectors, cables and strain relief. M12 circular connectors and sealed cable glands are common for power, sensor and communication interfaces. Contact plating, gasket design and locking mechanisms must match the target IP class and expected number of mating cycles. Where pigtail cables are used, strain relief clamps on the housing or mounting bracket should carry all mechanical loads, leaving solder joints and PCB pads free from tensile and bending stress caused by wind, ice and accidental pulls on the cable runs.

PCB safety, coating and surge paths. Printed circuit boards close to medium- or low-voltage conductors must respect clearance and creepage distances between high-voltage domains and low-voltage logic, using slots, keep-out zones and dedicated isolation components. Conformal coating improves tolerance to humidity and pollution but must be applied with attention to vents, connectors and rework practices. Surge and lightning energy should be routed through short, wide copper paths from terminal blocks and protectors to a ground or bonding point, keeping high currents away from sensitive AFEs and MCUs. Detailed SPD grading, earthing and shielding remain topics for the dedicated surge protection page.

Reference designs & IC mapping

Grid IoT nodes can be built around highly integrated MCU plus LPWAN SoCs or around discrete combinations of AFEs, microcontrollers, LPWAN modules and energy-harvesting PMICs. Integrated devices minimise footprint and component count for focused LoRa or cellular designs, while modular architectures allow reuse of the same sensing front-end across several radio options and vendors. The table below highlights representative vendors and part families for each role rather than attempting to be exhaustive or to replace detailed parametric selection tools.

All-in-one MCU + LPWAN SoCs. Devices such as STM32WL55 from STMicroelectronics, nRF9160 from Nordic Semiconductor or EFR32 sub-GHz SoCs from Silicon Labs integrate a low-power ARM core with on-chip radio and security features. These parts suit compact nodes where a single RF band or protocol dominates and where simplified routing and reduced BOM are more important than maximum flexibility in mixing AFEs and modems from different suppliers.

Discrete AFE + MCU + LPWAN + PMIC combinations. Many grid nodes use precision AFEs from Analog Devices or Texas Instruments, paired with ultra-low-power MCUs from ST, Microchip or TI and LPWAN modules from vendors such as u-blox, Quectel or Murata. Energy-harvesting PMICs from TI or Analog Devices manage solar, CT or thermal inputs and charge supercapacitors and batteries. This modular approach supports multiple node variants, regional RF differences and long-term second-source strategies while keeping the mechanical footprint and firmware architecture consistent.

Application mini-stories

Pole-top icing-monitoring node: vibration, tilt and LoRaWAN

On overhead lines in cold regions, icing increases conductor weight, changes sag and alters the vibration signature of spans and fittings. A pole-top icing-monitoring node combines a low-power MEMS accelerometer, a basic tilt or inclination measurement and one or more temperature sensors to track how a representative span behaves over a winter season. The sensing front-end is kept simple, favouring digital MEMS and modest ADC requirements, but must respect high-voltage clearance and creepage around the line hardware.

Power is harvested from a small solar panel mounted near the cross-arm, backed by a supercapacitor and a secondary cell managed by an energy-harvesting PMIC. This combination allows short bursts of high-current radio activity even on overcast days, while keeping average consumption in the hundreds of microwatts over 24 hours. LoRaWAN in a private utility network suits clusters of such nodes along a corridor, using one or two gateways near substations or line intersections rather than relying on variable cellular coverage in remote terrain.

Icing node data feed into higher-level telemetry and asset-health functions as one more stream of evidence about span condition. Trend analysis can correlate tilt and vibration events with temperature and weather data, while simple thresholds generate maintenance tickets when sag or vibration crosses defined limits. The Telemetry & Asset Health page describes how such node streams interact with fault records, switching events and regional asset scores, but the node itself only needs to deliver accurate, timestamped measurements and well-structured alarms.

• Power concept: small pole-top solar panel + supercapacitor + secondary cell, designed for no planned battery replacement over a decade.
• Wireless link: LoRaWAN Class A in a private, substation-centred network, with long reporting intervals and event-driven alarms.
• Sensing front-end: low-power MEMS accelerometer, inclination measurement and temperature channels, sharing a duty-cycled AFE and ADC.
• Data usage: sag and vibration trends inform patrol prioritisation and icing risk maps, while threshold crossings raise targeted work orders.

Ring Main Unit cabinet node: temperature, door status and NB-IoT

Urban and suburban Ring Main Units (RMUs) sit in pavements, basements and compact kiosks, often with limited ventilation and mixed environmental exposure. A simple cabinet node monitors several temperature points on busbar joints and cable terminations, the internal air temperature and the cabinet door status, and accepts a digital input from a partial-discharge indicator relay or optical coupler. The sensor mix aims to detect overheating, moisture-related issues and unauthorised access without duplicating full power-quality or PD-analysis equipment.

This node typically draws its main power from the RMU auxiliary supply, with a small backup battery or supercapacitor providing short-term ride-through and orderly shutdown during outages. The cabinet locations are widely dispersed but mostly within cellular coverage, making NB-IoT a natural choice for the backhaul. Temperature and humidity trends can be batched and reported every 15 minutes or hour, while door-open events and partial-discharge alarms trigger immediate uplinks with tighter latency requirements.

In the wider telemetry and asset-health context, RMU node data help rank cabinets by stress and risk. Frequent temperature excursions, repeated door openings and recurring PD indications can move a unit up the inspection queue, drive targeted maintenance and support planning for refurbishment or replacement. The cabinet node itself remains simple: it logs local events, applies basic thresholds and delivers clean datasets and alarms to downstream systems.

• Power concept: auxiliary DC supply inside the RMU, with local energy storage for short-duration backup and graceful shutdown.
• Wireless link: NB-IoT connection to the utility’s DMS or analytics platform, optimised for small, infrequent payloads.
• Sensing front-end: multi-channel temperature inputs, door switch and one or more digital alarm inputs from PD indicators or other relays.
• Data usage: temperature and event histories feed cabinet health scores and inspection priorities in Telemetry & Asset Health workflows.

Design checklist & engineering inputs

This checklist helps specify a grid IoT node so that suppliers or internal teams can size AFEs, MCUs, LPWAN links and power trees correctly. Each item describes a decision that influences component selection and mechanical layout. The examples in the table show typical values for overhead line and cabinet deployments; real projects should replace them with site-specific constraints and targets.

• Deployment environment and climate zone (overhead line pole, RMU, indoor substation; temperature, humidity and salt-mist exposure).
• Power concept and maintenance interval (battery only, energy harvesting, hybrid supply; years without battery replacement or manual intervention).
• Required analogue and digital sensor channels, including expected accuracy for key measurements.
• Sampling and reporting patterns (local sample intervals, uplink period, alarm latency tolerance and buffering behaviour during outages).
• Preferred LPWAN technology (LoRaWAN, NB-IoT, LTE-M or combinations) and security requirements such as encryption, secure boot and key storage.
• Mechanical envelope, mounting style, connector family, IP rating and any applicable standards or test levels for vibration, surge and environmental stress.

Grid IoT node – frequently asked questions

These FAQs turn the design trade-offs on this page into quick decision helpers: whether a custom node is justified, how to size the power budget, how to choose LoRaWAN versus NB-IoT, and how to deal with mechanics, security and lifecycle cost planning for thousands of deployed devices.