Antarctica is not a military theatre. But it imposes operating conditions that would challenge any military sensor deployment: extreme cold, no solar energy for months, no personnel present for maintenance, no reliable communications, and a consequence of failure — irreplaceable data lost, a research base potentially damaged — that focuses the mind sharply on reliability.
In 2024, ThingsLog deployed its LPMDL-1105 data loggers at the Bulgarian Antarctic Base St. Kliment Ohridski on Livingston Island. The deployment was designed to monitor indoor and outdoor environmental conditions through the Antarctic polar winter — the period from April to November when the base is unoccupied, the ambient temperature falls below −30 °C, and there is no sunlight to charge any solar panels.
The system operated autonomously for the entire polar winter without human intervention, battery replacement, or maintenance. The data it collected became the subject of a peer-reviewed publication at the IEEE CompSysTech 2025 conference.
This case study describes the deployment in detail and draws the explicit parallels to military unattended sensor network architecture — parallels that were not incidental but architectural.
The Challenge: Autonomous Winter Monitoring of an Unmanned Base
The operational context
The Bulgarian Antarctic Base St. Kliment Ohridski on Livingston Island is a seasonal research facility. Scientists and support staff occupy the base during the Antarctic summer (approximately November to April). During the polar winter — roughly April to November — the base is completely unmanned, all building services are shut down for safety, and there is no permanent presence on the island.
During this unoccupied period, several monitoring requirements exist:
- Structural integrity monitoring — has a storm damaged the buildings? Is there snow loading that poses a structural risk?
- Environmental baseline — what are the outdoor temperature, humidity, and pressure conditions during the winter? How do they relate to ice dynamics and local climate?
- Indoor vs outdoor comparison — do the buildings maintain thermal mass that buffers outdoor extremes? Or do interior temperatures track outdoor conditions directly?
- Anomaly detection — has anything changed that maintenance crews need to address before the next summer season?
All of these requirements must be met with no personnel present, no reliable mains power, no continuous internet connection, and no ability to replace a failed sensor node until the next summer season — a gap of seven months.
The power problem
This is the same power problem that defines military unattended ground sensor design:
– You cannot connect to mains power
– You cannot recharge solar batteries through polar winter (no sunlight)
– You cannot send maintenance crews to replace batteries
– Therefore, the system must operate on a fixed energy budget for the full deployment period
For the Antarctic deployment, the energy budget was determined by the available primary battery capacity and the required operating lifetime: approximately seven months of autonomous operation.
The communications problem
This is the same communications problem that defines military IoBT in denied environments:
– There is no cellular network
– The base has a Starlink satellite terminal for internet — but it consumes significant power, and power at the base during winter is provided only by a small bank of batteries charged before departure
– Therefore, the Starlink terminal cannot be left on continuously
The communications architecture must accommodate intermittent, scheduled connectivity — exactly as a military sensor in a denied or degraded communications environment must do.
The Solution: ThingsLog LPMDL-1105 with Duty-Cycled LoRa and Starlink
Hardware: LPMDL-1105 data logger
The LPMDL-1105 is ThingsLog’s ultra-low-power data logger designed for long-duration unattended monitoring applications. Key specifications relevant to the Antarctic deployment:
| Parameter | Value |
|---|---|
| Operating temperature | −40 °C to +70 °C |
| Radio | LoRa (Semtech SX1276-compatible), configurable SF and bandwidth |
| Protocol | ThingsLog Open LoRa Protocol |
| Sensor interfaces | 4× universal analog/digital sensor ports |
| Local storage | Non-volatile flash memory — configurable buffer depth |
| Power source | AA lithium primary cells (LiFeS₂ chemistry for low-temperature performance) |
| Sleep current | <10 µA (complete system including sensors) |
| Sampling interval | Configurable 1 second to 24 hours |
| Firmware | Event-driven, duty-cycled, store-and-forward |
Sensor configuration for Antarctica:
– Outdoor air temperature and relative humidity (4 sensors total across outdoor locations)
– Indoor temperature and humidity (inside selected base buildings)
– Barometric pressure (outdoor)
Each logger acquired readings from four sensors every 15 minutes, storing each reading to local flash memory — a total of 96 readings per day per sensor channel.
Architecture: Duty-cycled LoRa gateway and Starlink backhaul
The key architectural innovation of the Antarctic deployment was the coordinated duty cycling of three systems:
┌────────────────────────────────────────────────────────────┐
│ LPMDL-1105 sensor nodes │
│ • Acquire 4 sensors every 15 minutes │
│ • Buffer 96 readings (24 hours) in local flash │
│ • Sleep between measurements (< 10 µA) │
│ • Wake during daily communication window │
└─────────────────────────┬──────────────────────────────────┘
│ LoRa uplink (daily window)
┌─────────────────────────▼──────────────────────────────────┐
│ LoRa gateway (battery-powered) │
│ • Powered on once per day for ~60 minutes │
│ • Receives accumulated data from all sensor nodes │
│ • Forwards data to Starlink terminal via Ethernet │
│ • Returns to powered-off state │
└─────────────────────────┬──────────────────────────────────┘
│ Internet uplink (daily window)
┌─────────────────────────▼──────────────────────────────────┐
│ Starlink terminal (battery-powered) │
│ • Powered on once per day for ~30 minutes │
│ • Uploads sensor data to ThingsLog cloud platform │
│ • Returns to powered-off state │
└─────────────────────────┬──────────────────────────────────┘
│ HTTPS API
┌─────────────────────────▼──────────────────────────────────┐
│ ThingsLog cloud platform │
│ • Stores and visualises received sensor data │
│ • Alerts on missing daily uploads (node health check) │
└────────────────────────────────────────────────────────────┘
Why this architecture works:
– Each sensor node spends 99%+ of its time in deep sleep, consuming <10 µA
– The gateway and Starlink terminal are only powered for ~90 minutes per day — reducing their contribution to the daily energy budget by approximately 16×
– Even if a single daily communication window fails (due to poor satellite visibility or interference), the sensor nodes have already buffered the data locally — it will be uploaded in the next successful window
– The system can tolerate communication outages of many days without losing data, as long as the local flash buffer does not overflow
Radio configuration: Fixed SF8, ADR disabled
A critical decision in the Antarctic deployment was the LoRa radio configuration. Standard LoRaWAN uses Adaptive Data Rate (ADR) to automatically select the optimal spreading factor for the current link conditions — typically choosing a lower spreading factor (higher data rate, shorter time-on-air) when the link is strong.
For the Antarctic deployment, ADR was disabled and spreading factor was fixed at SF8. The reason: the sensors needed to transmit 96 readings per sensor channel in a single uplink payload during the daily communication window. At lower spreading factors (SF7), the maximum LoRaWAN payload size of 242 bytes would be insufficient to carry this data volume. At SF8 with the ThingsLog Open LoRa Protocol’s efficient binary encoding, the entire 24-hour data packet fits within the available payload capacity.
This is a direct analogue of the military consideration of optimising LoRa configuration for data volume rather than purely link reliability — a system design choice that reflects operational requirements rather than defaulting to protocol defaults.
What the Polar Winter Taught Us
The Antarctic deployment operated through the polar winter of 2024, producing continuous monitoring data that provided several insights of direct relevance to military and long-duration sensing applications:
Thermal performance at polar temperatures
The LPMDL-1105 operated continuously at outdoor ambient temperatures reaching −28 °C. The AA LiFeS₂ (lithium iron disulphide) cells maintained adequate voltage throughout, confirming their suitability down to −40 °C as rated. Commercial alkaline cells would have failed at these temperatures — their internal resistance rises dramatically below 0 °C, and they may freeze below −18 °C.
Military lesson: Primary battery chemistry selection is critical for cold-weather unattended sensing. LiFeS₂ (AA “Ultimate Lithium” cells) and LiSO₂ (military BB-2590 chemistry) are the appropriate choices for sub-zero deployments.
Indoor vs outdoor thermal dynamics
One of the scientific contributions of the deployment was the first published comparison of indoor and outdoor conditions at the base during polar winter. Contrary to expectations, building interiors did not maintain significantly warmer temperatures than the exterior during sustained cold periods — thermal mass effects were limited by the absence of any internal heat source.
Military lesson: Electronic equipment housed in nominally sheltered structures (tents, shipping containers, bunkers) cannot be assumed to be in a temperature-moderated environment during sustained cold weather operations. Worst-case operating temperature should be assumed equal to ambient outdoor temperature for equipment qualification purposes.
Store-and-forward resilience
On several occasions during the winter, the daily communication window failed — poor atmospheric conditions, occasional Starlink service interruptions, or gateway power issues prevented successful data upload. In every case, the sensor nodes continued to buffer data locally, and the missed data was successfully uploaded when the next successful communication window occurred.
The system sustained data gaps of up to 72 hours without permanent data loss, because the local flash buffer had been sized to hold several days of readings. No data recorded during the Antarctic winter was lost due to communication failures.
Military lesson: In any denied or degraded communications environment, local buffering with configurable buffer depth is not optional — it is the fundamental reliability mechanism. The sizing of the local buffer should be calculated from the expected maximum communications outage duration, with safety margin.
Gateway power budget validation
The battery bank powering the LoRa gateway and Starlink terminal was sized by the deployment team to last the full polar winter. The actual energy consumption was within 15% of the pre-deployment estimate — validating the power budget model and confirming that the duty-cycled operation architecture delivers predictable energy consumption.
Military lesson: Accurate power budget modelling is achievable if the duty cycle is well-defined and the active-state power consumption of each subsystem is measured on the actual hardware. Relying on datasheet specifications without hardware measurement introduces significant uncertainty.
Military Parallels: Where This Architecture Maps Directly
The Antarctic deployment was not designed as a military demonstrator. But the architectural decisions made in response to the extreme environment of Livingston Island map almost exactly onto the requirements of military unattended sensor systems:
| Antarctic constraint | Military equivalent |
|---|---|
| No mains power, no solar in winter | Denied environment, no resupply |
| No maintenance access for 7 months | Unattended multi-year sensor lifetime requirement |
| Intermittent Starlink connectivity | Denied/degraded communications environment |
| Extreme cold (−28 °C outdoor) | Arctic theatre operations |
| No personnel present | Unattended ground sensor (UGS) deployment |
| Data must not be lost during comms outages | Tactical sensor data continuity requirement |
| System must not generate waste heat that could be detected | Low thermal signature (relevant for military covert sensors) |
The Antarctic deployment is, in all structural essentials, an unattended polar sensor network operating under military-equivalent constraints. The LPMDL-1105 passed its “operational test” without qualification testing under MIL-STD-810H — but having survived a polar winter, its environmental credentials are compelling.
Standards Alignment
The LPMDL-1105 is not currently qualified under full MIL-STD-810H or STANAG 4370. However, the Antarctic deployment provides empirical evidence of performance at conditions that align with several MIL-STD-810H test methods:
| MIL-STD-810H Method | Condition tested | Antarctic equivalent |
|---|---|---|
| Method 502.7 — Low Temperature | Storage −57 °C, Operation −40 °C | Operation at −28 °C outdoor, storage potentially lower |
| Method 507.6 — Humidity | 95% RH | Antarctic fog and condensation cycles |
| Method 514.8 — Vibration | Transport vibration profiles | Ship transport to Livingston Island |
| Method 509.7 — Salt Fog | 96 hours 5% NaCl | Marine salt spray during ship transport and coastal installation |
For defence-adjacent programmes that require formal MIL-STD-810H or STANAG 4370 qualification of LPMDL hardware, ThingsLog can engage with accredited test laboratories to complete the qualification programme. Contact us to discuss specific environmental requirements.
Publication Reference
The Antarctic deployment is documented in:
Milovanov, N. “Deployment of a Low-Power LoRa-Based Monitoring Network for Environmental and Building Condition Assessment in Antarctica,” in Proceedings of the International Conference on Computer Systems and Technologies (CompSysTech ’25), 2025.
A related publication on LoRa/LoRaWAN for challenging LPWAN deployments:
Milovanov, N. “Case study: LoRa/LoRaWAN as a suitable LPWAN choice for pressure and flow monitoring in Bulgarian water utility sector,” in CompSysTech ’24, pp. 173–179, 2024. DOI: 10.1145/3674912.3674945.
Working with ThingsLog on Defence-Adjacent Deployments
The Antarctic deployment validates ThingsLog’s LPMDL platform for the class of requirements that define military unattended sensor networks: extreme environment operation, multi-year battery life, local data buffering, duty-cycled communication, and operational resilience in the absence of human intervention.
If you are working on a deployment with similar characteristics — whether in civilian infrastructure protection, critical facility monitoring, or a defence-adjacent or dual-use sensing programme — contact ThingsLog to discuss how our hardware and protocol stack can be adapted to your specific requirements.
Series Navigation
- Why Low Power Matters in Military Operations
- Key Application Domains
- How Military Low-Power Electronics Are Built
- Protective Coatings for Military Electronics
- Military Electronics Standards
- IP Ratings and Ingress Protection
- Case Study: DARPA N-ZERO
- Case Study: LoRa Tactical Troop Tracking
- Case Study: ThingsLog LPMDL in Antarctica
- Case Study : Army CombatConnect
