We develop custom embedded firmware for microcontroller-based devices: from board bring-up to a product in production. Specialised in ultra-low-power BLE architectures with Zephyr RTOS on Nordic nRF52/nRF54, ESP32 and STM32, we write firmware that is robust, maintainable and documented — not a prototype that only works on the bench.
The device was meant to last a year and lasts three weeks. We review sleep management, the BLE radio profile, sensor timing and the transmission policy until consumption is back within the project budget.
Intermittent crashes, stack overflows, race conditions, a watchdog that keeps firing. We move the firmware onto a solid RTOS (Zephyr) with a clear task architecture, logging and structured debugging.
You have a board (your own or a third party's) and you need someone to bring it to life: bring-up, sensor drivers (I²C/SPI), bootloader, DFU, BLE integration and a companion app.
The prototype is fine, but you need production-ready firmware: error handling, OTA updates, provisioning, end-of-line testing and deliverable documentation.
BLE 5 advertising and connected, custom GATT, BLE Mesh. Consumption optimisation for years of autonomy on a coin cell. This is our home turf: the SailSensor ecosystem was born this way.
Multi-thread architectures on Zephyr, device tree, power management, integration with the Nordic nRF Connect SDK ecosystem.
Drivers for I²C/SPI/UART sensors, IMUs, pressure and environmental sensors. Sensor fusion (complementary filters, quaternions) directly on the microcontroller.
In-field firmware updates via OTA or DFU, secure bootloaders, partition management and rollback. Because a product needs updating after it is sold.
Unlike a software-only developer, we also look at the schematic and the PCB: if the problem is hardware we pinpoint it, instead of passing the blame back and forth. All services →
A bench with oscilloscope, logic analyzer and BLE sniffer: firmware is verified by measurement, not by gut feeling.
Every project under NDA from the first contact. Sources and intellectual property remain the client's.
SailSensor BoomAngle measures the boom angle to better than 0.5° with two IMUs and differential quaternion fusion, in BLE advertising and with years of autonomy. It is low-power embedded firmware applied to a hard physical problem — the kind of work we do.
Yes. We work both on our own boards and on third-party hardware: we do the bring-up, write or fix the drivers and bring the firmware to production-ready level. Because we also have PCB and schematic skills, if the blocker is hardware we pinpoint it instead of staying stuck.
For simple projects, bare-metal is perfectly fine. When task concurrency, BLE, power management and OTA updates grow, an RTOS like Zephyr makes the firmware more robust, testable and maintainable over time — and it is officially supported by Nordic through the nRF Connect SDK. We choose the approach based on the project, not on fashion.
It depends on the duty cycle, but with BLE advertising, aggressive sleep management and low-power sensors, reaching years of autonomy on a coin cell is realistic. This is exactly the starting point of our SailSensor products: consumption is designed in, not discovered at the end.
Yes. At the end of the project we deliver the source code, the technical documentation and the build instructions. The intellectual property of the firmware developed for the client belongs to the client. Everything is covered by NDA.
The studio is in Cavezzo, in the province of Modena (Emilia-Romagna, Italy), EU timezone. Most firmware work is already done remotely, with on-site meetings only when a project genuinely needs them — for hardware bring-up, for example.
Yes, we are set up for it: EU timezone, documentation and communication in English, NDA signed before any technical detail is shared, and deliverables (source code, schematics, test reports) handed over in full. For a company looking to outsource RTOS firmware development services rather than hire in-house, this gives a single accountable point of contact instead of a distributed team to manage.
Tell us about the device, the microcontroller (or the idea) and the consumption target. In the initial discovery, free of charge, we assess feasibility, risks and timelines, and propose a path with defined costs.
Start with a free discovery