The NewSpace revolution, and particularly the emergence of CubeSats and nanosatellites, has created new opportunities for embedded systems technologies. Modern small satellites increasingly rely on low-cost, low-power, and highly integrated components, including resource-constrained microcontrollers (MCUs), MEMS sensors, off-the-shelves imagers, software-defined radio technologies, and low-power communication transceivers such as LoRa® devices. These technologies enable the rapid development and deployment of affordable space missions while significantly reducing development cycles and operational costs.

The ThingSat project is an in-orbit LoRa gateway designed as an experimental platform for evaluating Low-Power Wide-Area Network (LPWAN) communications in space. The project investigates several emerging use cases, including Satellite Internet of Things (SatIoT), inter-satellite links (ISL), and the prototyping of innovative and affordable satellite services. Potential applications include Positioning, Navigation and Timing (PNT), ranging and localization services, Emergency and Warning Services (EWS), Search and Rescue (SaR), Automatic Identification System (AIS) reception, Direct-to-cell for the poor man (Meshcore and Meshtastic) and other distributed communication modes (SCHC …).

This presentation will provide an overview of the software architecture developed for ThingSat and its integration with the RIOT operating system. Particular attention will be given to the reuse and adaptation of key open-source middleware components and drivers required for space operation. The presented software stack includes:

– A mission engine scheduling SatCom experiments according to time, orbital position, and mission constraints.
– A space-adapted SX1303 LoRa gateway driver enabling multi-channel LoRa packet reception and transmission;
– Support for repeat and store-carry-and forward protocols for LoRaWAN, Meshcore and Meshtastic networks
– SUIT (Software Updates for the Internet of Things) for secure remote firmware over-the-air software updates of CubeSat subsystems through the spacecraft onboard communication buses, including CAN, RS-485, and I²C;
– TensorFlow Lite Micro (TFLM) for onboard artificial intelligence and edge inference, enabling real-time attitude anomaly and outlier detection from IMU sensor measurements;

The next generation of the firmware will progressively embrace the Rust programming language to improve memory safety, reliability, and long-term maintainability of critical onboard software. It will be built upon Ariel OS, a modern Rust-based embedded operating system designed for resource-constrained devices, while preserving interoperability with the existing RIOT OS ecosystem and software components. This evolution aims to leverage Rust’s strong safety guarantees to reduce common software vulnerabilities and increase the robustness of future CubeSat and NewSpace missions.

The presentation will also discuss the challenges of deploying terrestrial IoT technologies in the space environment, including radiation constraints, intermittent connectivity, power limitations, and long communication delays. Finally, lessons learned from the development and in-orbit operation of ThingSat will be shared, highlighting how open-source embedded software ecosystems can accelerate innovation in NewSpace missions and contribute to the emergence of next-generation satellite services.