Cybernetica enters third phase of ESA’s Minerva project with on-board spacecraft cybersecurity module
Cybernetica has begun the third phase of the Minerva project in partnership with the European Space Agency (ESA). Building on more than five years of collaborative research, this phase marks a pivotal transition: from ground-based machine learning exploration to a deployable, real-time cybersecurity system embedded directly into spacecraft.
The Minerva Dynamic Threat Detection System is a dedicated cybersecurity module designed to sit at the critical security boundary between a spacecraft’s communication system and its on-board computer (OBC). The project runs under ESA’s General Support Technology Programme (GSTP) and is active until 2029.
From research to hardware
Cybernetica’s partnership with ESA began in April 2020, when the agency selected the company to de-risk machine learning technology for cyber situational awareness. That first phase focused on developing a novel toolset — combining entropy-based structural modelling with unsupervised machine learning — capable of identifying suspicious network behaviour with less manual analyst workload.
The second phase developed those research outcomes into a functional software prototype, demonstrating the potential for automated behaviour analytics.
The third phase now brings that intelligence on board the spacecraft itself.
“Space is no longer a domain where isolation provides security. As spacecraft become nodes in interconnected networks, they inherit the same vulnerabilities as any networked system on Earth. Bringing real-time threat detection on board is an essential next step,” said Sander Valvas, Head of Cybersecurity at Cybernetica.
A dedicated on-board cybersecurity module
At the core of this phase is MINERVA-H, a plug-and-play hardware module that installs between the spacecraft’s telecommand/telemetry (TM/TC) receiver and its on-board computer. It supports multiple spacecraft bus protocols including Ethernet, SpaceWire, UART, and CAN, enabling integration across a wide range of spacecraft architectures.
For missions unable to accommodate additional hardware, the project explores development of MINERVA-S, a software-only option that runs directly on the existing OBC. Both options are paired with MINERVA-GSC, a ground segment companion platform that manages security policy updates, AI/ML model retraining, threat signature distribution, and forensic analytics.
The system employs hybrid threat detection: threat signatures catch known attack patterns, while an AI/ML behaviour analytics engine identifies anomalous command sequences. All detection, response, and logging occur in real time, before potentially hostile commands can affect mission-critical subsystems.
Why spacecraft cybersecurity is urgent
Spacecraft have historically been designed with reliability and radiation tolerance as the primary constraints. Cybersecurity was assumed to be addressed at the ground station. As space assets become increasingly interconnected, drawing on commercial cloud infrastructure, shared ground networks, and multi-operator command chains, the attack surface has expanded.
Threats have become multifaceted: rogue ground stations, command-and-control interference, laser interference and satellite blinding, on-board malware and more. A successful intrusion into a spacecraft’s OBC can be irreversible.
Valvas states that the number of cyber incidents targeting space infrastructure has grown significantly over the past five years: “What concerns us most is their sophistication: adversaries are specifically probing command-and-control channels. With Minerva, we are putting the security boundary directly on the spacecraft, before a hostile command can ever reach the on-board computer.”
Thus, Minerva addresses this by creating an independent security boundary on-board.
Source: Cybernetica