By 2030, there will be an estimated 100,000 satellites in space, and in most cases, their primary function is to send information back to Earth.

Advances in technology, however, mean a new generation of satellites can produce exponentially more data than their predecessors. This is causing the space equivalent of a traffic jam. In short, more data is being collected than can be sent back to Earth efficiently and on time.

Traffic bottlenecks are never a good thing, either on Earth or in space, and while those of us on the ground continue to fume at the delays, Dublin-based Setanta Space Systems is aiming to solve the problem for space users with a cutting-edge solution designed to ease the growing congestion.

“Basically, we’re addressing one of the most persistent limitations in modern spacecraft: the lack of high-performance, flexible computing at the edge,” explains company co-founder James Murphy.

“Satellites are generating increasing volumes of data and operating in more complex environments, yet much of today’s avionics still relies on legacy architectures that constrain autonomy and require heavy dependence on ground operations,” he adds.

Setanta’s approach is to put “intelligence” directly into space vehicles via a family of modular and scalable computing platforms that combine radiation-tolerant electronics with deployable AI.

This, in turn, enables spacecraft to process sensor data and make operational decisions in orbit rather than waiting for instructions from Earth.

“We believe spacecraft need to be smarter, and our goal is to provide the hardware and software foundations that enable true on-board autonomy. We want intelligence to be built into every mission from day one,” Murphy says.

Murphy’s PhD focused on AI for spacecraft and he has spent the last decade in industry, working closely with the European Space Agency.

He was the software architect behind the Viki video telemetry system flown on Ariane 5 and Ariane 6, and before setting up Setanta with co-founder Adam Taylor last year, he was AI engineering lead at Réaltra Space Systems. There, he was responsible for the development of flight-ready embedded software and AI systems for launch vehicles and satellites. Co-founder Taylor is an engineer with over 20 years’ experience in the space and defence sectors.

“Our core platform is called Danu [after the goddess of Irish mythology], an on-board computer built around what we refer to as our Space Tile architecture. This is a flexible architecture that allows mission designers to tailor processing power, AI acceleration, and radiation tolerance depending on need.

“Modular hardware allows computing subsystems to be integrated, upgraded or reconfigured without having to redesign an entire spacecraft avionics stack,” Murphy says.

Alongside the hardware, Setanta is also developing AI models for spacecraft health monitoring, Earth observation analytics and autonomous navigation. Together, these two solutions can make satellites more intelligent, resilient and capable of operating more independently, which is critical in challenging environments.

Years of research and space programme experience provided Murphy with the knowledge to identify and potentially solve the emerging congestion problem and he first began discussing the idea with Taylor over a pint about 18 months ago.

Setanta was formally established in July 2025 and began operating last December. When it moved into NovaUCD, formal development work, product planning, and early customer/commercial engagement began.

Employment at Setanta will rise to eight by the summer, with consultancy assignments and early angel investment funding the venture as the company prepares its products to the required level of technical readiness for flight.

“It’s hard to put an absolute figure on the investment so far, but it’s somewhere between €200,000–€300,000 in combined founder time, engineering effort, prototyping, and early-stage hardware production.

“This includes direct founder capital, reinvested contract revenues and the contribution of significant pre-existing intellectual property developed over several years in the space sector,” says Murphy, who describes what he does as “sitting at the intersection of advanced AI research and space-qualified hardware, with a strong focus on building practical, deployable systems for next-generation space missions”.

Setanta is an Enterprise Ireland (EI) client, but to date, the agency’s support has been more by way of advisory and contacts than money, as EI has a strong connection with the European Space Agency (ESA). Supplying the space industry might sound like a tall order for a start-up, but Murphy says the European space sector is very open to working with fledgling companies and SMEs. The ESA is also active in supporting smaller operators and collaborations, and partnerships are potentially fruitful avenues that start-ups and young businesses can pursue.

Setanta’s aspirations are international as the space industry operates across borders and components are typically integrated into global supply chains. That said, the company is focused more on Europe for now, with the US and beyond to follow. Likely customers for the company are satellite manufacturers, mission integrators and space technology companies developing small satellites and constellations, both established and emerging.

Setanta by no means has the market to itself, however, Murphy says its disruptive technology will pique interest because it rethinks how spacecraft computing is procured and evolved over the lifetime of a mission.

“Today, satellite operators often face a trade-off between high-performance commercial electronics [that are not radiation-resilient] and traditional space-grade systems that can be slow to adapt and expensive to iterate,” he says.

“This slows down innovation and limits experimentation in orbit. By introducing a platform-based approach, where computing capability can be configured, upgraded, and scaled without redesigning the entire avionics stack, Setanta is enabling a consistent architecture across multiple mission profiles and reducing development friction and shortening iteration cycles.”