Firetail Hardware
“Not a Tag”
Open reference hardware for animal tracking research
A complete, ultra-low-power development platform that integrates GPS, acceleration, cellular connectivity, and on-device machine learning — everything you need to prototype the next generation of wildlife tracking devices.
Designed and validated by Schäuffelhut Berger GmbH. Built to work with Firetail from day one.
The gap we're filling
Commercial tags
Sealed black boxes. You can't modify firmware, experiment with sensors, or run custom algorithms. Great for deployment, useless for development.
Arduino / DIY
Cheap but lacks sensor quality, power management, cellular connectivity, and the ultra-low-power design needed for real field deployments.
“Not a Tag”
Professional-grade hardware that's open for development. Real sensors, real power management, real cellular — and you control the firmware.
What you can do with it
Develop custom tracking firmware
Full access to all peripherals via standard ARM toolchains. JTAG/SWD debug, USB-C power, and development headers make iteration fast.
Run ML models on the edge
The dedicated co-processor with FPU/DSP runs TensorFlow Lite for Microcontrollers. Train behavioral models in Firetail with FireSOM, deploy them directly to the tag.
Profile power consumption
Jumper-configurable power rails let you isolate and measure each subsystem independently. Validate that your firmware achieves the power budget for field deployment.
Test cellular & GNSS in the field
LTE Cat-M1/NB-IoT for data upload, multi-constellation GNSS for positioning. Test real-world connectivity and satellite acquisition in your deployment environment.
Prototype new sensor integrations
SPI, I²C, UART, and GPIO headers exposed. Add your own sensors, actuators, or peripherals to the platform.
Feed data directly into Firetail
Collected data integrates with Firetail for visualization and analysis. The complete pipeline from hardware to publication-ready visuals.
FireSOM on the edge
Train behavioral classification models in Firetail. Export to TensorFlow Lite. Flash to the tag. The device itself classifies behavior in real-time — hunting, resting, feeding, flying — and adapts its sampling strategy accordingly.
No other animal tracking platform offers this workflow.
Technical specifications
| Main MCU | STM32WL55 dual-core (Cortex-M4 + M0+) |
| Integrated radio | Sub-GHz (FSK/GFSK/BPSK) |
| Co-processor | STM32L4Q5CG (ARM Cortex-M4, up to 120 MHz, FPU/DSP for ML inference) |
| Cellular | LTE Cat-M1 / NB-IoT (Quectel BG77) |
| GNSS | GPS, GLONASS, Galileo, BeiDou |
| Motion sensor | 9-axis IMU (accelerometer, gyroscope, magnetometer) |
| Temperature | Ambient temperature sensor |
| Power | Solar harvesting (MPPT), LiPo battery, USB-C |
| Sleep current | 120 nA |
| Storage | 2 Gbit SPI flash for local data buffering |
| Debug | JTAG/SWD, debug LEDs, push buttons, power profiling jumpers |
| Form factor | Development board with exposed headers and test points |
Based on a bachelor thesis at the University of Applied Sciences Munich, 2025. Supervised by Friedrich Schäuffelhut, Schäuffelhut Berger GmbH.
Who is this for?
Research labs
Building custom tracking solutions for specific species or environments. Need a validated platform instead of starting from scratch.
Tag developers
Prototyping next-generation devices. Use "Not a Tag" as a reference design with proven power characteristics.
PhD students
Working on firmware, embedded ML, or hardware research in movement ecology. A ready-made platform to build your thesis on.
Conservation tech programs
Developing custom monitoring solutions for specific field deployments where commercial tags don't fit.
Interested in “Not a Tag”?
“Not a Tag” is a research and development platform and is not available for sale. If you'd like to learn more about the platform, discuss a research collaboration, or hear about future plans, get in touch.
Get more info