The technology stack behind our products
We choose the right silicon, language and protocol for each problem, and we have worked hands-on with all of the tools below.
Controllers
From ultra-low-power MCUs to edge AI processors.
Microcontrollers
- ARM Cortex-M0/M3/M4/M7
- STM32
- ESP32 / ESP32-S3 / C3
- Nordic nRF52 / nRF53 / nRF54L
- NXP Kinetis / LPC
- Microchip PIC / AVR
- TI MSP430
Edge processors
- NVIDIA Jetson
- Raspberry Pi / CM4
- NXP i.MX
- Rockchip
- Embedded Linux SBCs
Programming Languages
Firmware, edge, cloud and apps, in the language that fits.
Embedded
- C
- C++
- Embedded Rust
- MicroPython
- Assembly
Edge, AI & cloud
- Python
- Node.js / TypeScript
- PHP
- SQL
- Go
Apps
- Dart / Flutter
- Kotlin
- JavaScript
Communication Protocols
On-board buses, industrial field buses and application protocols.
On-board
- UART
- SPI
- I²C
- I²S
- 1-Wire
- USB
- SWD / JTAG
Industrial
- Modbus RTU / TCP
- CAN / CAN FD
- RS-485 / RS-232
- OPC UA
- 4–20 mA
Application
- MQTT
- HTTP / REST
- WebSocket
- CoAP
- UDP
Networking Technologies
Short range, long range and cellular, chosen for range, power and cost.
Short range
- Bluetooth Low Energy
- Wi-Fi
- Zigbee
- Thread
- UWB
Long range & cellular
- LoRa / LoRaWAN
- 4G LTE
- LTE-M / NB-IoT
- GSM / GPRS
- GNSS
Wired
- Ethernet
- PoE
- Fibre
Drones
Electronics, firmware and payloads for UAVs used in inspection and monitoring.
Flight & control
- Flight controller firmware
- PX4 / ArduPilot
- MAVLink
- GNSS / RTK positioning
- Telemetry over 4G, LoRa and RF links
Payloads
- RGB and thermal cameras
- LiDAR
- Gas and environmental sensors
- Gimbal control
Applications
- Aerial infrastructure inspection
- Onboard edge AI
- Geo-tagged imaging and mapping
Robotics
Embedded control, sensing and autonomy for robots that work in the field and on the shop floor.
Control
- Motor control: BLDC, stepper, servo
- Field-oriented control (FOC)
- Encoders and IMU sensor fusion
- Real-time firmware
Autonomy & perception
- ROS 2
- SLAM and navigation
- LiDAR and depth cameras
- Machine vision
Applications
- AGVs and mobile robots
- Inspection robots
- Robotic arm integration
Semiconductor and technology experience
Chips and modules we have designed with, and the connectivity standards we build on.
AI for every kind of device data
What we build for each data type, with examples of where it is used.
Capability
Anomaly detection, forecasting, predictive maintenance, sensor drift correction and device self-diagnostics.
Example
Spotting a slowly falling voltage trend on overhead railway lines days before it trips a section.
Capability
On-device keyword spotting, voice commands, speech-to-text and text-to-speech.
Example
Hands-free voice control for field technicians, working offline on a low-power device.
Capability
Acoustic event detection, machine-sound fault classification and noise monitoring.
Example
Detecting bearing wear in pumps and motors from changes in their sound signature.
Capability
Object detection, defect inspection, OCR and reading of analogue gauges and meters.
Example
Digitising old dial meters with a camera module, without replacing the meter.
Capability
Real-time tracking, intrusion and obstacle detection, activity recognition at the edge.
Example
Spotting rockfall or obstructions on a track and alerting the control room within seconds.
Capability
NLP and LLM assistants over logs, maintenance tickets, manuals and reports.
Example
Summarising fault history and suggesting next actions to maintenance engineers.
AI tooling
- TensorFlow / TFLite Micro
- PyTorch
- ONNX Runtime
- Edge Impulse
- OpenCV
- YOLO
- TensorRT
- scikit-learn
Need help choosing a stack?
We will help you weigh up silicon, radio and cloud options for cost, power and scale.