Figure 1. Hewlett Packard Enterprise Spaceborne Computer-2 (left) onboard the International Space Station (right).

Figure 2. During the first-of-its-kind in space demonstration, the SaraniaSat team identified a single flying airplane in the vicinity of Dubai.

Building a Cloud in Space

Making this onboard cloud-native computing demonstration possible required an entirely new generation of tools. BBI and Netnod contributed Dacreo Apto, a Kubernetes-compatible “spacestack” that transforms satellites into cloud-native platforms. By embedding execution-monitoring agents, hardened OS primitives, and seamless BigOps/DevOps workflows, Dacreo Apto allows resource-constrained satellites to behave like resilient, self-governing compute nodes. For mission teams, this means the ability to iterate, deploy, and recover applications in orbit just as easily as they would in a terrestrial cloud environment. Transmitting 155 gigabits per second of raw H-Cubed™ data to Earth is impossible with the current LEO satellite communication bandwidth limitations, and when combined with the latency of terrestrial processing undermines daily, critical decision making in agriculture and many other fields that depend on timely, actionable, geospatial information.

ENVI® in Orbit

For SaraniaSat, another critical partner was NV5. The in-space demonstration showcased the portability of ENVI Deep Learning, a geospatial analytics platform with roots dating back to 1977. ENVI Deep Learning automates the detection, classification, and change monitoring of objects and features in imagery. Originally built to handle hyperspectral data, ENVI has matured into a versatile tool for delivering actionable intelligence from complex multispectral and hyperspectral datasets. In this test, ENVI proved its ability to run effectively on HPE’s Spaceborne computer-2, bringing decades of geospatial image analysis expertise into the newest computer frontier: the cloud in space. From raw inputs to actionable intelligence, ENVI has enabled SaraniaSat to push boundaries not only in Defense-based applications such as Moving Target Identification (Fig. 2) but also in enabling, global, data-driven, precision agriculture.

As Dr. Thomas (Tom) George, founder of SaraniaSat, explained, “ENVI was designed for hyperspectral data, and its sophistication makes it perfect for handling the complexity of H-Cubed™ data.” From detecting moving targets in defense to enabling precision agriculture, ENVI remains central to SarniaSat’s vision.

From Agriculture to Analytics: SaraniaSat’s Origins

That vision traces back to SaraniaSat’s founding in 2016. Tom launched SaraniaSat to address a problem that has eluded Earth observation for decades: enabling daily, global, data-driven agriculture. This data is critical for the entire Agricultural Ecosystem including Farmers, Agricultural Companies, Crop Insurance and Crop Finance Companies, Crop Futures Markets, and Food Supply Chain Companies. Despite more than 50 years of both government and private Satellite Missions, beginning with NASA’s Landsat-1 mission in 1972, the agriculture ecosystem still lacks timely, reliable, and actionable data at a global scale.

Tom describes the challenge as a “Don Quixote quest.” He points to two barriers: complexity and timeliness. Without H-Cubed™ data, farmers cannot make confident decisions without risking their already razor-thin margins. And without near-real-time access, data loses its value across the typical 100-150 day agricultural crop cycle, where weather, pests, disease, and stress conditions can change by the day.

For Tom, this makes the case clear: only by combining complex data with high-performance edge computing onboard satellites can agriculture and other industries overcome these daily, decision-making challenges.

Catalysts for a New Industry

The implications go far beyond agriculture. Tom believes in-space cloud computing could reshape geospatial intelligence much like GPS and the internet changed daily life. “The GPS inventors in the 1960s never imagined Siri giving us directions to the grocery store today,” he notes. “I am particularly excited by novel applications enabled by in-space, high performance edge computing of remote sensing big data that we can’t even imagine today!”

The July demonstration proved that such a future is possible. With partners like NV5, BBI, HPE, and Netnod, SaraniaSat is blazing a trail for a new generation of geospatial solutions. “Lives, missions, and billions of dollars depend on timely information,” Tom says. “The answer is clearly to move compute power closer to the source: onboard satellites.”

Despite the rollercoaster journey of entrepreneurship (hours of sheer terror punctuated by moments of boredom), Tom remains motivated by the people and the mission. “I’m humbled and grateful to work with such talented, collaborative partners,” he reflects. “What keeps me motivated is the possibility of leaving the world better than I found it and solving problems that very few thought were solvable.”

Looking Forward

For Tom and his collaborators, this demonstration is more than a milestone: it’s a catalyst. By proving that cloud-native analytics and AI can thrive in orbit, they have opened doors to applications that will transform defense, intelligence, agriculture and beyond.

Have a bold idea or a complex problem that seems unsolvable? Connect with the experts at NV5 (GeospatialInfo@NV5.com). We’ll help you turn your vision into reality.

REFERENCES

1. SaraniaSat’s Cloud‑Native AI/ML in Orbit Demonstration
2. Dacreo Apto a Kubernetes-compatible spacestack developed by BruhnBruhn Innovation
3. https://www.saraniasat.com/