By Nick David, Editorial Lead, SatNews
For most of Earth observation’s history, you bought imagery one wavelength at a time. Optical from one provider, radar from another, thermal from a third, hyperspectral from a fourth if you had a research budget and patience. Each sensor was its own program with its own ground segment, ordering interface, tasking lag, and analyst workflow. The integration happened in the customer’s organisation, after the data left the constellation. Day 2 of SmallSat Europe 2026 was the meeting where that architecture started to crack in public, and the cracks ran through four separate sessions.
Fusion replaces single-sensor
The strategic frame came from the Multi-Physics EO Stack panel, moderated by Dr. James Crawford, CSO of Privateer Space. Crawford put the technical thesis on the table at the open. The future of Earth observation is not a better optical satellite or a better SAR satellite. The future is fused multi-physics data, captured close in time across the constellation, with a common tasking layer that lets a customer ask a question once and receive an answer back from whichever sensor combination delivers the highest-signal observation.
“I work for the architect office, and we’re trying to look at the overall ecosystem in space and getting away from individual satellite missions. We’re really trying to push this working together, making sure that the systems are interoperable, that there is a common tasking, and that the inter-satellite links are there.” Vanessa Keuck, EO System Architect, European Space Agency
ESA’s job, Keuck said, is to make EU member states able to procure into a multi-physics fusion layer instead of each one building its own individual satellite mission. The upcoming procurement instrument she pointed to is an Invitation to Tender ESA expects to launch addressed at commercially enabling small-satellite missions for environmental policy.
Matthew Jenkins of Vantor, the rebrand of Maxar Intelligence after its take-private last year, made the parallel commercial case. Vantor has changed its public identity from “satellite imaging provider” to “software-enabled spatial intelligence company” precisely because the customer value has moved from the pixel to the integration. Vantor markets four pillars now. Tasking via the WorldView and recently launched Legion satellites. 2D and 3D Vivid Imagery global mosaics. Mission Enable products including the Raptor alternate PNT product and Century site monitoring. And the Tensor Globe AI platform that ties the others together.
The Cosine, KSAT, and Simera Sense voices on the panel each pivoted to the same conclusion from a different angle. Cosine’s Marco Esposito described an edge-processed multi-sensor ISR system that flags early-warning events from thermal IR and tips other modalities into action. KSAT’s representative described the Norwegian institutional Earth-observation business as a fully sovereign service delivered without owning any satellites, by fusing data from many sources. Simera Sense’s Thys Cronje, Chief Commercial Officer, described a customer that combined a thermal camera with a multispectral camera on a single platform to read refinery activity in Iran before and after strikes. Different companies. Same architectural conclusion. The product is the fused intelligence delivered into a customer workflow, not the imagery.
Optical mesh is a manufacturing problem now
Inter-satellite optical links are no longer experimental. They are a manufacturing problem. The Optical Mesh panel was moderated by Dr. Victor Aguero of Cambrian Works and was structured exactly to test how far away European production is from delivering hundreds or thousands of terminals to orbit.
The participants laid out the architecture. Mynaric (Lubos Fedora), Tesat Spacecom (Daniel Tröndle), and Cailabs (Dr. Jean-François Morizur) on the optical side. Kepler Communications (Michael Dowell) as the mission operator running commercial optical backhaul. FSO Instruments (Gus van der Feltz) and the ESTOL/SDA interop framework defining the standards. Cailabs CEO Morizur framed his company’s position plainly on stage: the 200-person France-and-US business runs optical ground stations on three continents and does not build space terminals, it connects to them.
The structural argument across the panel was simple. Launch costs have fallen far enough that small countries, companies, and even individuals can put satellites on orbit. All those satellites need to exchange information. The available RF spectrum does not scale that far. Optical mesh becomes the only viable architecture above a certain network density. The European optical-terminal manufacturers and ground-station operators on the panel were each ramping toward mass-production facilities, with the FSO Instruments speaker explicitly framing their Delft build as “ready for mass production.” The procurement signal from ESA’s IRIS² and from national defence ministries is now strong enough to underwrite the capex.
The integration layer is the product
The Making the Most of Commercial Imagery panel, moderated by Christy Monaco, Chief Operating Officer of the Open Geospatial Consortium, took the integration thesis into the defence buyer’s seat. Tero Vauraste of Kuva Space, Patrick Butler of Sidus Space, Thys Cronje of Simera Sense, and Jonathan Debilde of Blacksky kept circling the same observation. The defence buyer is not asking for resolution. The defence buyer is asking for a workflow.
The procurement-pace problem followed naturally. Defence organisations move slowly. Commercial constellations move quickly. The integration layer that absorbs the speed mismatch is the actual product. Without it, the constellation operator sells imagery and the defence buyer struggles to operationalise it. With it, the constellation operator sells decisions, and the defence buyer integrates them into existing field workflows.
Anthony Baker, CEO of SatVu, delivered the tech brief that put the operational example in front of the panel. SatVu’s thermal imagery is 3.5-metre resolution, high enough to see what is happening inside structures, and Baker put the file-size number on the slide. Less than fifty megabits per tile, sometimes less. Small enough to transmit through messaging apps to a front-line user without breaking the bandwidth budget.
The verdict from Day 2
The value in Earth observation has moved from owning a sensor to integrating across sensors. From building a satellite to building a tasking platform. From owning the data to delivering the intelligence into a workflow. The constellations that win the next five years are not the ones with the highest pixel counts. They are the ones with the most defensible integration story, the cleanest API into a customer workflow, and the lowest-friction path from question to answer.
The pixel war is over. The integration war is what comes next.
Key Takeaway
Earth observation’s competitive frontier moved off the sensor in 2026 and onto the tasking and integration layer. ESA is procuring into that layer with a new ITT for commercially enabling environmental missions. Vantor restructured around it. The optical-mesh suppliers are ramping serial production to support it. The defence buyer is asking for it. The constellation operator that delivers a clean workflow wins the next five-year procurement cycle.
About the Author
A storyteller at heart, Nick David covers space policy, satellite markets, defense, and the technologies reshaping how humanity operates beyond Earth. With a background in creative direction, brand strategy, and editorial storytelling, he brings a modern lens to complex subjects and a relentless curiosity about what comes next.