June 23, 2026 – Ubotica Technologies, the Irish space technology firm pioneering Orbital AI, has today announced an $11 million funding round to accelerate deployment of its AI-powered intelligence platform, a breakthrough maritime security service powered by real-time decision-making satellites in the Earth’s orbit.

The round was led by Act Venture Capital and Greencode Ventures, with participation from existing investor Atlantic Bridge.

As industries plan for orbital compute, Ubotica has already proven AI at the orbital edge. Its Orbital AI technology has run hundreds of thousands of AI inferences in orbit, deployed more than 30 Earth observation models on board satellites across multiple missions, and achieved a 100% mission success rate. Working with NASA’s Jet Propulsion Laboratory and the European Space Agency, it has delivered a series of world firsts, including the first spacecraft to autonomously identify a target ahead and reorient itself to capture it.

The investment will accelerate the commercial rollout of Ubotica’s Live Maritime Intelligence platform, which enables governments and maritime agencies around the world to detect emerging threats across vast maritime territories in real time.

The funding comes as nations face growing pressure to protect critical maritime infrastructure, including undersea communications cables, offshore energy assets, pipelines and strategic shipping routes, while responding to increasingly complex activity such as shadow fleets, dark vessels, sanctions evasion and potential sabotage.

Over the past 12 months, Ubotica has seen significant demand for its Live Maritime Intelligence, driven by its ability to enhance maritime surveillance across exclusive economic zones that are often far larger than a nation’s landmass.

Live Maritime Intelligence works by continuously identifying areas of elevated maritime risk, helping operators prioritise surveillance resources.
Rather than relying on fixed satellite collection schedules and post-processing on Earth, the platform combines Orbital AI with autonomous ground-to-orbit tasking to dynamically focus available sensing assets on emerging activity, providing earlier warning and faster operational intelligence back down to Earth.

Monitoring risk from space

While most Earth observation satellites still capture imagery before transmitting it to Earth for processing, Ubotica has fundamentally changed that model.

Orbital AI significantly reduces response times on Earth while making satellite networks far more efficient, by moving intelligence closer to where data is captured.

Ubotica is among the first companies to have proven Orbital AI in operational space missions, amid growing interest and investment around the space sector, in-orbit computing and AI infrastructure in space.

Orbital AI pioneers

“Ubotica has spent years pioneering Orbital AI, and we have applied that knowledge to one of the hardest security challenges on Earth: protecting vast maritime zones and critical offshore infrastructure,” said Fintan Buckley, CEO of Ubotica Technologies. “Live Maritime Intelligence predicts where risk is emerging, tasks the right satellites and sensors, and delivers decision-grade intelligence in minutes, giving security teams the speed and efficiency they need to act. This investment allows us to bring LMI to market at scale.”

“Ubotica has been running inference in orbit for years and transforms static Earth Observation into live intelligence. Nations can’t afford to wait hours for imagery while a vessel goes dark over critical infrastructure. Live Maritime Intelligence proves what on-board intelligence is worth: it turns Earth Observation from a camera into a real-time sentinel, starting with Europe’s waters,” said Andrew O’Neill, Act Venture Capital

“Resilience is a key topic for Europe in the coming years. Securing critical infrastructure in harbors and off-shore is crucial. Ubotica’s edge AI technology enables satellite constellations to detect threats and anomalies early, anywhere across vast ocean areas, creating both a strategic security capability and a compelling commercial opportunity,” said Terhi Vapola, Greencode Ventures.

Dublin and Arlington, 18 May 2026 – Ubotica Technologies, a pioneer in space-based AI, has today announced a partnership agreement with NOVI Space Inc. (“NOVI”), a global leader in orbit edge computing and next-generation satellite infrastructure. The partnership will deliver decision-grade, real-time intelligence directly from Earth’s orbit.

The collaboration integrates Ubotica’s SPACE:AI platform with NOVI’s GENIE smart-satellite constellation and platform, enabling Earth observation data to be processed directly onboard satellites, supporting faster decision-making in time-critical domains such as maritime monitoring.

Traditional Earth observation systems collect first and think later. Raw imagery is downlinked, queued for ground processing, and delivered hours or days after capture. This is far too slow for threats that move in minutes.

Onboard AI eliminates that delay. Through the partnership, Ubotica will deploy and operate AI models directly in orbit through the GENIE multi-sensor satellite platform powered by NOVI’s SP240 On-Board Computer, enabling rapid detection, classification and decision-making.

SPACE:AI generates intelligence within 90 seconds of acquisition and delivers actionable insights in near real-time, without transmitting the underlying imagery. The result is dramatically lower latency, lower bandwidth cost and higher intelligence value per pass.

Onboard processing fundamentally changes what a satellite does. Instead of acting as a passive collector of pixels, each satellite becomes an intelligent agent: capturing, analysing and delivering insights without ground processing in the loop. SPACE:AI has been proven across multiple models and over 100,000 in-orbit inferences. In a single observation of Singapore port, it processed hundreds of vessels and detected those operating dark in under two minutes, with insights on the ground immediately dispatched.

The economics shift accordingly. Bandwidth costs fall sharply. Latency collapses from days to minutes. A single pass returns intelligence rather than imagery, ready for operational use the moment it reaches the ground.

Dr. Aubrey Dunne, co-founder and CTO at Ubotica, said:
“Our collaboration with NOVI brings more AI-enabled Earth Observation capacity into orbit. By combining SPACE:AI with GENIE’s onboard compute, we’re shifting satellites from data collectors to intelligent agents, delivering insights in minutes rather than days. That capability underpins our Live Maritime Intelligence service and unlocks new operational models for time-critical surveillance.”

Michael Bartholomeusz, CEO of NOVI Space, added:
“AI at the edge is not a future concept, it’s happening now. Partnering with Ubotica allows us to demonstrate the full potential of GENIE: delivering real-time intelligence from orbit, not just data. This is a fundamental shift in how space-based systems create value.”

Pasadena / Dublin / London, April 14 2026 – Ubotica Technologies and Open Cosmos today announced an agreement with NASA’s Jet Propulsion Laboratory (JPL) to collaborate on the Flight Demonstration of Federated Autonomous MEasurement (FAME), under the Advanced Information Systems Technology programme in NASA’s Earth Science Technology Office.

FAME aims to link more than 50 spacecraft from a wide range of operators in the largest autonomous satellite operations test ever attempted, demonstrating a fundamentally new way to observe Earth.

Ubotica’s SPACE:AI platform and two Open Cosmos spacecraft, the currently operational Hammer and Accenture-1, are named as AI satellite assets in the programme.

From Observing to Knowing and Acting

Today’s Earth observation satellites – optical, radar, infrared – collect vast amounts of data across multiple modalities and return it to the ground for human analysts to interpret hours or days later. While the satellites have powerful instruments, they have no understanding of what they are seeing.

FAME builds on a new generation of satellites that do understand the data they are collecting. When Ubotica’s onboard AI processes data in orbit and detects an event of interest, such as ships at sea that have switched off their tracking systems, a wildfire, or a volcanic eruption, it acts immediately: takes a closer look, generates an alert, tasks a follow-on spacecraft. No analyst in the middle. No ground station delay.

FAME federates that capability across an entire constellation. One satellite sees something. The network of satellites responds. Follow-on spacecraft reorient and acquire targeted observations within minutes. The loop closes autonomously, replacing passive data collection with active, autonomous, federated intelligence across a constellation that can coordinate as an intelligent system.

The multiyear flight demonstration of FAME is expected to begin with an initial set of six spacecraft in summer 2026.

Built on Two Space AI Breakthroughs

FAME builds on two innovations demonstrated in orbit for the first time aboard Hammer.

Real-Time Identification

Ubotica’s SPACE:AI platform processes imagery onboard, identifying events of interest in seconds rather than waiting for ground-based analysis.

Autonomous Action through Dynamic Targeting

Dynamic Targeting, a technology developed by NASA JPL and first successfully demonstrated in orbit by NASA JPL, Ubotica and Open Cosmos in July 2025, autonomously reorients the spacecraft and captures high-resolution confirmation imagery in just over 60 seconds, with no ground station involvement.

Together, these two capabilities produce more valuable data, and the insights are available much quicker than with a conventional pass.

FAME is the next stage: taking these proven single-satellite capabilities and making them work across a federated constellation.

Leadership Perspective

Fintan Buckley, CEO at Ubotica said: “Dynamic Targeting showed what a single satellite with onboard AI can achieve. FAME shows what happens when that capability is coordinated across a network. Our contribution is the intelligence inside the Ubotica nodes: detecting what matters, processing it in orbit, and passing the signal to whatever asset can act on it fastest. That is how you close the loop at a speed that is actually useful.”

Rafel Jordá Siquier, CEO of Open Cosmos said: “When Open Cosmos builds and designs satellites, we always have two key goals in mind: to understand and connect the world. Understanding the world requires a layer of context that is accessible in real-time, so that data can be acted upon without delay. By adding the scale of a full network to this capability, FAME opens a new realm of possibility for what Earth observation technology can achieve.”

Programme Timeline

Year one focuses on maturing flight capabilities and executing initial onboard AI and rapid notification tests across the core constellation. Years two and three scale to the full 50+ spacecraft network, processing thousands of automated alerts and executing hundreds of autonomous on-orbit tasking commands across assets from multiple operators and agencies.

Operational Context

The SPACE:AI platform is already in operational deployment in Ubotica’s Live Maritime Intelligence service, providing predictive maritime domain awareness across European waters. The same onboard AI that detects dark vessels and monitors critical underwater infrastructure powers Ubotica’s contribution to FAME.

3 December, 2025 – Ubotica Technologies, NASA’s Jet Propulsion Laboratory and Open Cosmos Honored with the 2025 SpaceNews Icon Award for Space AI Partnership

Ubotica Technologies, NASA’s Jet Propulsion Laboratory and Open Cosmos have been honored with the 2025 SpaceNews Icon Award for Space AI Partnership in recognition of their joint work on Dynamic Targeting, a technology that uses artificial intelligence to help spacecraft decide, autonomously and within seconds, where to make science observations from orbit.

The award was presented at the 8th SpaceNews Icon Awards ceremony, held on December 2, 2025, at the Hopkins Bloomberg Center in Washington, D.C., where leading figures from across the global space community gathered to celebrate this year’s honorees.

The award recognizes the team’s efforts to harness artificial intelligence for advancing space operations and capabilities.

SpaceNews Icon Awards

Now in its eighth year, the SpaceNews Icon Awards honor individuals, missions and organizations whose innovation and leadership are shaping the global space industry.

Presented annually by SpaceNews, a leading publication covering the business, policy and technology of space since 1989, the awards spotlight excellence in categories such as Commercial Space, Civil Space, Innovative Technology, Sustainability and Space AI Partnership.

A collaborative breakthrough

The Dynamic Targeting demonstration was developed at JPL in collaboration with Open Cosmos and Ubotica Technologies. The test took place on Hammer, a satellite designed, built and operated by Open Cosmos, and is used for Ubotica’s CogniSAT-6 mission. The satellite’s onboard autonomy payload, featuring a commercially available AI processor, was designed and developed by Ubotica.

Funded by NASA’s Earth Science Technology Office and developed at JPL, Dynamic Targeting allows a satellite to autonomously look ahead on its orbital path, analyze imagery captured by the spacecraft’s camera, and decide within seconds where to aim its sensors, capturing the most valuable data.

The technology demonstrates how onboard AI can enable Earth-observing satellites to process and analyze look-ahead imagery in real time to determine and immediately capture the best follow-up image, a process that takes less than 90 seconds without human intervention — fully autonomous.

The first flight test demonstrated Dynamic Targeting’s ability to avoid imaging clouds, a common challenge for Earth-observing spacecraft. By analyzing imagery from a look-ahead view roughly 300 miles (500 kilometers) in front of its path, the spacecraft can distinguish between clouds and clear sky and choose where to capture an image.

This approach yields a higher proportion of usable science data. It enables satellites to be smarter about what they image, ensuring they are not storing, processing and downlinking imagery that is unusable to scientists.

Aubrey Dunne, Chief Technology Officer at Ubotica, said: “Receiving the SpaceNews Icon Award is a huge honor for the entire team. It reflects years of work developing space-qualifying autonomous systems that enable satellites to see, think and act for themselves. This recognition affirms that artificial intelligence is now becoming a core capability for space missions.”

Steve Chien, Fellow in Artificial Intelligence at JPL, said: “This Dynamic Targeting demonstration is a significant step forward in shifting decision-making from mission controllers on Earth to the spacecraft. The spacecraft can now, on its own, acquire and analyze data in tens of seconds and immediately focus on what is most important. Dynamic Targeting is a powerful tool with the potential to revolutionize space missions — enabling dynamic measurements, avoiding clouds for atmospheric retrieval missions, or tracking plumes at a comet.”

Jordi Barrera Ars, Chief Technology Officer at Open Cosmos, said: “We are immensely proud that the efforts of the team to make satellites more autonomous — being used more efficiently to increase

Ubotica is proud to be named a laureate of the European Space Agency’s Rising Stars Award, recognising high-potential European companies shaping the future of space. The award highlights the growing importance of autonomous satellite technologies and reflects Ubotica’s progress in
delivering real-time intelligence through our SPACE:AI platform.

SPACE:AI enables satellites to Think. See. Act. by combining on-ground orchestration, on-orbit AI processing, and a modular suite of SPACE:AI Agents such as Dark Vessel Detection and Dynamic Targeting. By extracting insight directly on orbit, SPACE:AI delivers Live Earth Intelligence without relying on slow downlink cycles, transforming how operators access time-critical information for security, maritime awareness, disaster response, and environmental monitoring.

The ESA Rising Stars recognition strengthens Ubotica’s position within Europe’s mission and commercialisation ecosystem and supports deeper collaboration on programmes where real-time, autonomous satellite operations are essential.

Over the past couple of years, Ubotica has advanced SPACE:AI through key milestones, including the launch of CogniSAT-6, the start of major European programmes such as MESEO, and continued research with collaborators including the NASA Jet Propulsion Laboratory. We have also expanded our maritime intelligence capabilities through ongoing work in dark vessel activity and ship-to-ship transfer detection, demonstrating the operational value and maturity of SPACE:AI.

Ubotica thanks ESA for this recognition, and acknowledges the support of our team, partners, and customers. We look forward to expanding the SPACE:AI ecosystem and continuing to advance intelligent, autonomous satellite capabilities.

The Ubotica team are still at the Space Tech Expo Europe, in Bremen until tomorrow, drop by booth D56 for a chat.

In the last three weeks, Europe’s maritime policy has shifted significantly. Three developments — new sanctions, expanded inspection powers, and a reminder of chokepoint fragility — signal a move from static enforcement to dynamic, behaviour-based monitoring of the seas.

On October 19, a draft EU Maritime Declaration proposed empowering European authorities to inspect high-risk vessels through cooperation with flag states. The declaration targets ships suspected of disabling AIS, engaging in covert ship-to-ship (STS) transfers, or operating under ambiguous registry status.

Just days later, on October 23, the European Commission unveiled its nineteenth sanctions package, expanding restrictions to 557 vessels linked to Russia and associated networks. The measures extend beyond simple asset freezes; the Commission urged flag-state coordination and data sharing to track high-risk vessel activity in real time.

Then, on October 28, a sanctioned tanker briefly disrupted traffic in the Suez Canal, halting dozens of ships and exposing how European energy and trade routes remain vulnerable to incidents far beyond EU waters.

Together, these events reveal a new reality: Europe’s maritime security risks no longer stem solely from who owns the ships, but from how those ships behave.

From Naming Ships to Knowing Their Behaviour

For years, Europe’s maritime enforcement strategy has focused on static lists, sanctioned entities, flagged vessels, and access restrictions. While essential, this model presumes that risk is fixed and identifiable. In reality, illicit maritime activity adapts faster than the frameworks built to control it.

Vessels engaged in covert trade no longer simply change ownership or flags; they manipulate behaviour. Ships disable AIS to avoid detection, loiter near anchorage zones at night, or conduct unreported STS transfers in unmonitored waters. Each action erodes the reliability of Europe’s surveillance systems and regulatory assumptions.

The shift now underway — from identity-based compliance to behaviour-based intelligence — represents an evolution in how Europe must see its seas.

Why Compliance Alone No Longer Works

The EU’s growing blacklist is a powerful diplomatic tool, but compliance mechanisms alone cannot secure Europe’s maritime domain. Even vessels formally denied entry to EU ports can affect European interests by disrupting chokepoints, distorting market flows, or causing environmental damage.

The Suez Canal incident is illustrative. A single sanctioned tanker created hours of disruption across one of the world’s most vital arteries, a reminder that Europe’s energy and logistics security depends on visibility across a global network — not just inside European waters.

In practice, this means surveillance strategies must extend to key corridors such as the Suez, Gibraltar, and the Danish Straits — zones that connect Europe to the rest of the world’s supply lines. The challenge is not just to monitor them, but to predict and pre-empt anomalies before they escalate.

Behaviour as the New Compliance Metric

Behaviour is harder to falsify than identity. Repeated AIS gaps near restricted zones, unusual loitering patterns, and recurrent course deviations under low visibility — these are all indicators of potential illegal activity that no sanctions registry can capture.

For policymakers, this means developing behavioural triggers: quantifiable indicators that prompt inspection, enforcement, or targeted observation. Rather than reacting to known offenders, authorities can anticipate risk based on how vessels act.

This approach aligns with the broader intelligence community’s shift toward predictive analytics and multi-modal data fusion, blending satellite imagery, AIS data, and contextual metadata to detect anomalies in real time.

And yet, Europe’s ability to act on that intelligence remains limited by a fundamental constraint: time.

The Latency Gap — and the Case for SPACE:AI

Surveillance latency — the delay between observation, analysis, and response — is where enforcement efforts often fail. Conventional satellite workflows rely on downlinking raw imagery to the ground, processing it hours or days later, and only then drawing conclusions. By that time, dark vessels have already moved on.

Bridging this latency gap requires on-orbit intelligence: satellites that can detect, analyse, and react autonomously at the edge.

This is where SPACE:AI, Ubotica’s end-to-end framework for Live Maritime Intelligence, becomes pivotal. By processing data directly on orbit, SPACE:AI transforms satellites into intelligent agents that can identify behavioural anomalies in real time — such as unreported STS transfers, AIS gaps, or formation patterns suggesting covert logistics.

Instead of waiting for imagery to be processed on the ground, SPACE:AI delivers actionable insight within minutes, providing authorities the temporal advantage they currently lack.

From Observation to Action

The EU’s evolving maritime doctrine demands systems that do more than observe — they must interpret and act. Through on-orbit AI models like Dynamic Targeting (for autonomous re-tasking), SPACE:AI enables this transformation.

These capabilities allow Europe to move beyond static surveillance into a new era of autonomous maritime intelligence, where satellites can dynamically prioritise high-risk zones, alert operators to suspicious behaviour, and cue further observation automatically.

It’s not simply about adding more satellites — it’s about making them smarter.

Conclusion

Europe’s maritime frontier is changing. Blacklists and sanctions will continue to expand, but the true frontier lies in detecting how ships operate, not just who operates them.

Behaviour-based intelligence, powered by frameworks like SPACE:AI, marks the next phase of maritime governance — transforming enforcement from reactive to proactive, from static documentation to living insight.

But behaviour cannot be understood through a single lens. It requires the fusion of multiple data sources — optical and SAR imagery, AIS signals, RF emissions, and contextual environmental data — analysed together to form a coherent, real-time maritime picture.

This is the foundation of connected intelligence: a system where satellites collaborate, data layers reinforce one another, and AI agents continuously learn from patterns across the global maritime domain.

The defining question for Europe’s next decade at sea will not be:
Which ships are sanctioned?

But rather:
Which ships are behaving as though they should be?

Because in an age of adaptive evasion, behaviour is the new identity — and multi-source, real-time intelligence is the only defence.

Join Ubotica at Space Tech Expo 2025: Advancing Live Earth Intelligence with SPACE:AI and Dynamic Targeting

Meet the team and experience SPACE:AI, Ubotica’s end-to-end platform that enables satellites to think, see, and act with advanced on-orbit AI. SPACE:AI transforms Earth Observation satellites into autonomous systems capable of delivering real-time, actionable insights — reducing latency, optimising bandwidth, and maximising mission efficiency.

We’ll also be showcasing Dynamic Targeting, a SPACE:AI Agent that autonomously detects, prioritizes, and tracks emerging events of interest — coordinating observations across a constellation to capture the right scene at the right moment. From maritime anomalies to environmental change, Dynamic Targeting accelerates decision-making and tasking-to-insight.

Book a meeting (click here) with the Ubotica team to discuss onboard processing and autonomous satellite operations tailored to your smallsat program.

More information on Space Tech Expo 2025:

Space Tech Expo Europe returns to Bremen, Germany, 18–20 November 2025, bringing together the continent’s leading space innovators. This year’s event features 10,000+ attendees, 950+ exhibitors, and 200+ speakers — Europe’s largest B2B space industry gathering. You’ll find Ubotica in Hall 4, Stand D56, near the Smallsats Conference area.

Conference themes include space exploration, sustainability, downstream services and applications, cyber security, onboard connectivity, and next-generation satellite networks, with speakers from ESA, Airbus, the German Space Agency (DLR), and more.

Registration (free)
Exhibitor list
Smallsats conference

Ubotica’s North Star

At Ubotica, our north star is simple: make satellites intelligent so the world can see, predict, and act in time. Running powerful AI on the spacecraft is how we get there.

When a vessel goes dark near critical subsea infrastructure, every minute matters. A SPACE:AI-enabled satellite detects the vessel, processes imagery on-board, and alerts operators in near real time—because the thinking happens in orbit, not days later on the ground.

But scaling this capability hits two walls: models become too large for space deployment, and labels are too expensive to generate. Our latest research tackles both—shrinking models so they run fast at low power on flight-class hardware, and training without human labels by learning from a larger model’s predictions.

The Dual Challenge

Size & Power: Foundation model ViTs like Prithvi and TerraMind are superb—but at ~86M and 103M parameters, they’re impractical for many satellites. Our student model (MobileNetV2+UNet) is ~3.73M parameters—~23–28× smaller and built for on-board use.

Label Scarcity: High-quality EO labels (flood boundaries, crop classes, vessel types) are costly and slow to produce, especially for rare events and new regions. Our question: can a small model learn from a big model’s outputs instead of human labels—and still generalise?

Our Solution: Teacher-Supervised Learning

We leveraged knowledge distillation—essentially, having a large “teacher” model train a small “student” model. Our student architecture (MobileNetV2+UNet with just 3.73 million parameters) is 23–28× smaller than the teacher models.

But we pushed further. Instead of just using the teacher to supplement labeled data, we asked: What if the teacher’s predictions could replace human labels entirely? If this works, we eliminate the annotation bottleneck for many tasks. The teacher becomes a “labeling oracle” that enables rapid model development without expensive manual annotation campaigns.

Testing the Hypothesis

We tested on two real-world Earth observation challenges:

• Sen1Floods11: Flood Detection — A binary classification task (water vs. not-water) with 446 labeled images. Critically, it includes an out-of-distribution test dataset (Bolivia dataset)—a completely different geographic region that tests whether models truly understand floods or just memorized training locations.
• Multi-Temporal Crop Classification — A complex 14-class problem using temporal satellite imagery across the continental United States. This tests whether teacher supervision scales to nuanced, multi-class scenarios.

Two Training Strategies

• Strategy A: Blended Learning — The student trains on both real ground truth labels and the teacher’s final-layer outputs, with a parameter controlling the balance.
• Strategy B: Pure Teacher Supervision — The student trains on the teacher’s predictions as pseudo-labels, in place of the ground truth. No human annotations. This directly tests whether we can bypass manual labeling entirely.

Results: Strong Performance, Especially for Generalization

Flood Detection: Excellent Results

• Prithvi as teacher: Standard test: 0.873 Dice (vs. 0.892 with real labels); Bolivia test: 0.845 (vs. 0.853 with real labels).
• TerraMind as teacher: Standard test: 0.899 (vs. 0.914 with real labels); Bolivia test: 0.866 (vs. 0.886 with real labels).

The teacher-supervised student achieved 95–98% of baseline performance and generalized remarkably well to the unseen Bolivia region.

Crop Classification: Good with Room for Improvement

• Prithvi as teacher: Test with real labels: 0.435 Dice score; Test with teacher predictions: 0.375 (about 86% of baseline).

Binary classification is more forgiving than 14-way classification. With more classes, teacher errors compound during student training. However, 0.375 without any human labels is still a viable starting point for many applications.

The Efficiency Payoff: Space-Ready Performance

Our student models run efficiently across satellite-suitable edge AI hardware:

• Snapdragon DSP: 637 FPS at 6.44W
• NVIDIA Orin Nano: 132 FPS at 5.43W
• MemryX MX3: 114 FPS at 4.15W
• Myriad X (heritage platform): 12 FPS at 3.11W

Even on power-constrained hardware, we achieve real-time performance. Knowledge distillation makes cutting-edge AI practical for space.

Real-World Impact on Live Maritime Intelligence

This research directly informs our LMI service:

• Faster fielding: Fine-tune a foundation model on the ground, then distill to a tiny flight model—no large annotation campaign required.
• Geographic transfer: The strong Bolivia results validate our maritime surveillance approach. A model trained on one region generalizes to new regions—critical for global infrastructure monitoring.
• Continuous improvement: As foundation models improve, we distill that knowledge to onboard models, creating an upgrade pathway for existing satellites.
• Lower barriers: Organizations can develop custom EO applications without investing heavily in annotation infrastructure.

Key Insights

• Teacher quality is paramount. Student performance ceiling is set by teacher accuracy.
• Task complexity matters. Teacher supervision excels for tasks with clear decision boundaries.
• Unlocks EO’s killer app. On-orbit inference compresses decision lead time from minutes to seconds.

The Road Ahead

We’re exploring more complex knowledge distillation approaches that better address multi-class problems and segmentation tasks with imbalanced datasets.

Conclusion: Foundation Models Meet Operational Reality

The AI revolution in Earth observation risks being Earthbound if we can’t deploy capabilities where they matter most—onboard the satellites capturing data.

Our research shows that knowledge distillation with teacher supervision removes two major barriers: model size and labeling costs. For binary and low-class-count problems, teacher-supervised students match traditionally trained models while generalizing to new regions. For complex tasks, there’s a gap, but results are encouraging enough to guide real deployments.

At Ubotica, this research directly informs how we build SPACE:AI capabilities today. Every insight about efficient training, improved generalization, and energy savings translates to more capable satellites delivering Live Maritime Intelligence when and where it’s needed.

With onboard AI powered by efficiently distilled models, satellites can watch, understand, and respond in real-time. When a dark vessel approaches critical infrastructure, our satellites know—immediately. That’s the future we’re building.

Ubotica Technologies has been named a finalist in the 2025 SpaceNews ICON Awards, recognised in the Space AI Partnership category alongside NASA’s Jet Propulsion Laboratory (JPL) and Open Cosmos for their collaborative work on Dynamic Targeting — an AI capability that transforms Earth observation from passive imaging to real-time Earth intelligence.

The SpaceNews ICON Awards celebrate organisations driving innovation and impact across the global space sector. The Space AI Partnership category recognises efforts to harness artificial intelligence for advancing space operations and capabilities.

Transforming Space into an Active Intelligence Network

Dynamic Targeting represents a paradigm shift in how satellites observe and respond to events on Earth. Instead of pre-scheduled imaging tasks, it enables satellites to think and act autonomously — detecting changes, assessing relevance, and triggering new observations in real time.

By bringing decision-making to orbit, this AI capability reduces latency, increases mission efficiency, and allows satellites to respond instantly to dynamic phenomena — from illegal maritime activity to natural disasters.

This achievement exemplifies the power of AI partnerships to accelerate innovation and unlock new levels of satellite autonomy.

Recognising Leadership in SPACE:AI

For Ubotica, this recognition affirms the company’s position as a global leader in SPACE:AI — the end-to-end AI platform that empowers satellites to think, see, and act in orbit. Dynamic Targeting is an evolving SPACE:AI capability that showcases how onboard intelligence can anticipate change — allowing satellites to look ahead, identify potential triggers in their path, and coordinate new observations in real time.

“This recognition underscores what’s possible when AI, autonomy, and collaboration converge in orbit,” said Fintan Buckley, CEO of Ubotica Technologies. “It marks another milestone in the evolution of intelligent satellite operations.”

About the SpaceNews ICON Awards

The annual SpaceNews ICON Awards highlight innovators and leaders shaping the future of the space industry. Finalists in the Space AI Partnership category also include Booz Allen Hamilton (with HPE & Meta) for deploying the first generative AI in space, and Mission Control Space Services with Spire for advancing onboard autonomy and decision-making.

Ubotica’s recognition underscores the company’s continued leadership in developing AI solutions that make space systems more autonomous, responsive, and intelligent — driving the next era of real-time space infrastructure.

A technology called Dynamic Targeting could enable spacecraft to decide, autonomously and within seconds, where to best make science observations from orbit.

In a recent test, NASA showed how artificial intelligence-based technology could help orbiting spacecraft provide more targeted and valuable science data. The technology enabled an Earth-observing satellite for the first time to look ahead along its orbital path, rapidly process and analyze imagery with onboard AI, and determine where to point an instrument. The whole process took less than 90 seconds, without any human involvement.

Called Dynamic Targeting, the concept has been in development for more than a decade at NASA’s Jet Propulsion Laboratory in Southern California. The first of a series of flight tests occurred aboard a commercial satellite in mid-July. The goal: to show the potential of Dynamic Targeting to enable orbiters to improve ground imaging by avoiding clouds and also to autonomously hunt for specific, short-lived phenomena like wildfires, volcanic eruptions, and rare storms.

“The idea is to make the spacecraft act more like a human: Instead of just seeing data, it’s thinking about what the data shows and how to respond,” says Steve Chien, a technical fellow in AI at JPL and principal investigator for the Dynamic Targeting project. “When a human sees a picture of trees burning, they understand it may indicate a forest fire, not just a collection of red and orange pixels. We’re trying to make the spacecraft have the ability to say, ‘That’s a fire,’ and then focus its sensors on the fire.”

Avoiding Clouds for Better Science

This first flight test for Dynamic Targeting wasn’t hunting specific phenomena like fires — that will come later. Instead, the point was avoiding an omnipresent phenomenon: clouds.

Most science instruments on orbiting spacecraft look down at whatever is beneath them. However, for Earth-observing satellites with optical sensors, clouds can get in the way as much as two-thirds of the time, blocking views of the surface. To overcome this, Dynamic Targeting looks 300 miles (500 kilometers) ahead and has the ability to distinguish between clouds and clear sky. If the scene is clear, the spacecraft images the surface when passing overhead. If it’s cloudy, the spacecraft cancels the imaging activity to save data storage for another target.

“If you can be smart about what you’re taking pictures of, then you only image the ground and skip the clouds. That way, you’re not storing, processing, and downloading all this imagery researchers really can’t use,” said Ben Smith of JPL, an associate with NASA’s Earth Science Technology Office, which funds the Dynamic Targeting work. “This technology will help scientists get a much higher proportion of usable data.”

How Dynamic Targeting Works

The testing is taking place on CogniSAT-6, a briefcase-size CubeSat that launched in March 2024. The satellite — designed, built, and operated by Open Cosmos — hosts a payload designed and developed by Ubotica featuring a commercially available AI processor. While working with Ubotica in 2022, Chien’s team conducted tests aboard the International Space Station running algorithms similar to those in Dynamic Targeting on the same type of processor. The results showed the combination could work for space-based remote sensing.

Since CogniSAT-6 lacks an imager dedicated to looking ahead, the spacecraft tilts forward 40 to 50 degrees to point its optical sensor, a camera that sees both visible and near-infrared light. Once look-ahead imagery has been acquired, Dynamic Targeting’s advanced algorithm, trained to identify clouds, analyzes it. Based on that analysis, the Dynamic Targeting planning software determines where to point the sensor for cloud-free views. Meanwhile, the satellite tilts back toward nadir (looking directly below the spacecraft) and snaps the planned imagery, capturing only the ground.

This all takes place in 60 to 90 seconds, depending on the original look-ahead angle, as the spacecraft speeds in low Earth orbit at nearly 17,000 mph (7.5 kilometers per second).

What’s Next

With the cloud-avoidance capability now proven, the next test will be hunting for storms and severe weather — essentially targeting clouds instead of avoiding them. Another test will be to search for thermal anomalies like wildfires and volcanic eruptions. The JPL team developed unique algorithms for each application.

“This initial deployment of Dynamic Targeting is a hugely important step,” Chien said. “The end goal is operational use on a science mission, making for a very agile instrument taking novel measurements.”

There are multiple visions for how that could happen — possibly even on spacecraft exploring the solar system. In fact, Chien and his JPL colleagues drew some inspiration for their Dynamic Targeting work from another project they had also worked on: using data from ESA’s (the European Space Agency’s) Rosetta orbiter to demonstrate the feasibility of autonomously detecting and imaging plumes emitted by comet 67P/Churyumov-Gerasimenko.

On Earth, adapting Dynamic Targeting for use with radar could allow scientists to study dangerous extreme winter weather events called deep convective ice storms, which are too rare and short-lived to closely observe with existing technologies. Specialized algorithms would identify these dense storm formations with a satellite’s look-ahead instrument. Then a powerful, focused radar would pivot to keep the ice clouds in view, “staring” at them as the spacecraft speeds by overhead and gathers a bounty of data over six to eight minutes.

Some ideas involve using Dynamic Targeting on multiple spacecraft: The results of onboard image analysis from a leading satellite could be rapidly communicated to a trailing satellite, which could be tasked with targeting specific phenomena. The data could even be fed to a constellation of dozens of orbiting spacecraft. Chien is leading a test of that concept, called Federated Autonomous MEasurement, beginning later this year.

This is a NASA-JPL Press Release.