GalaxEye: India’s Satellite Seeing Through Clouds & Its Impact

Ask a satellite to photograph a flooded district in Assam in July and you will often get a picture of cloud. Optical satellites work like cameras: they record sunlight bouncing off the ground. When a cyclone, a monsoon system or simply the night gets in the way, they have nothing to record. That is also, very often, exactly when someone needs the picture.

Synthetic aperture radar solves this differently. The satellite sends its own microwave pulses and times the echoes. Because it supplies its own illumination, it works at night, and because microwaves pass through cloud, it works in the monsoon. As the satellite moves, it combines echoes into what behaves like a very long antenna, which is where the word “synthetic aperture” comes from.

The catch is that radar pictures do not look like photographs. Calm water is black, buildings flare white, and fields show texture instead of colour. They take a trained analyst to read. So the trade-off has always been: optical is easier to understand; radar is more dependable. GalaxEye’s founders asked why you should have to choose.

Combining optical and radar data is not new. Analysts have long merged images from different satellites. The problem is that two satellites pass over at different times and angles, so the pictures never quite match: a ship has moved, a cloud has drifted, the flood has risen. GalaxEye’s approach is to put both sensors on one spacecraft, point them at the same spot at the same moment, and align the data onboard. It calls the product OptoSAR and the alignment SyncFusion.

On Drishti that meant an X-band SAR with a 3.5-metre deployable antenna and a seven-band multispectral imager covering panchromatic, red, green, blue, near-infrared, coastal blue and red edge, with an NVIDIA Jetson Orin module for onboard processing. The company claimed the fused product would carry up to three times more usable information than a single-sensor image. In September 2026 it won a US patent for the synchronisation system.

A new satellite does not start imaging the moment it separates from the rocket. First comes the launch and early orbit phase: establishing radio contact, unfolding antennas and solar panels, stabilising the spacecraft, checking every subsystem. GalaxEye had planned about eight weeks of commissioning and calibration before selling data. Drishti got through most of it. Then, in the final stage, an anomaly hit. Its signal became intermittent and then disappeared.

The company’s first explanation was space weather. An extreme geomagnetic storm, caused by a burst of solar activity, raised radiation levels, and GalaxEye said radiation effects likely damaged a critical onboard system. Radiation can flip bits in memory, latch up chips, or slowly degrade electronics; small satellites built partly from commercial parts are especially exposed. But this is the company’s preliminary finding. No independent investigation or detailed failure report has been published, and the exact mechanism is not public.

That makes Drishti neither a clean success nor a meaningless failure. It was a partial technical demonstration with a mission-ending fault. The spacecraft worked well enough to be tested; the payload never got its chance. For a first satellite from a start-up, that is a familiar outcome. The judgement that matters is whether the second one survives.

GalaxEye’s response to losing Drishti was to take more of the satellite in-house. In July it said it would accelerate moving supply chain, manufacturing and satellite development under its own control. On 10 August 2026 it bought StarOps, a Bengaluru spacecraft engineering firm with qualified satellite buses in the 50, 150 and 250 kg classes and engineers who had worked on the TeamIndus lunar lander. The logic is simple: a better radar does not help if the spacecraft around it does not survive.

A month later came the US patent for its optical-radar synchronisation, and a Reuters report that GalaxEye, having raised about 21 million dollars, was planning roughly 30 satellites over five years. That is a big step up from the “two satellites within 24 months” it had set in August, and from the 8 to 12 satellites in earlier plans. None has a launch date yet.

On 24 September 2026 the Technology Development Board signed on. Its support comes from the Research Development and Innovation Fund, the government’s vehicle for backing late-stage deep-tech development, as optionally convertible debentures rather than a grant. The target is ambitious: a multisensor satellite imaging finer than 0.5 metre, and a technology taken from TRL 6, a prototype shown in a relevant environment, to TRL 9, proven in actual operations.

Read closely, that TRL framing is also an official acknowledgement of Drishti’s outcome. The concept has been demonstrated in parts; it has not yet flown as a working system. The government is paying roughly a quarter of the cost of getting it there.

The real product is not the picture. A disaster agency does not want “SAR data”; it wants a list of cut-off roads. That is why GalaxEye pairs its sensors with onboard processing and AI, and why its pitch runs from image to data to insight to decision. It is also why one satellite is not enough: a single spacecraft passes over a given place only every few days. Useful monitoring needs a constellation.

GalaxEye is a product of India’s space reforms. IN-SPACe was set up in 2020 to authorise private space activity, and the Indian Space Policy 2023 opened every stage of the value chain, from building satellites to selling their data, to private companies. The government told Parliament that 105 authorisations had been issued to non-government entities as of 14 July 2026, and that 187 million dollars of private investment had been reported during 2026 alone.

Earth observation is where that shift is most visible. On 21 January 2026 IN-SPACe signed India’s first public-private Earth-observation constellation with a Pixxel-led consortium (with Dhruva Space, PierSight and SatSure): 12 satellites, including two X-band SAR satellites, and more than Rs 1,200 crore of private money by 2029. GalaxEye is not in that consortium. It is building its own fleet, betting on a sensor combination nobody else flies.

GalaxEye’s first satellite was meant to answer one question: can a private Indian company put an OptoSAR satellite into orbit? It did. The mission then raised a harder one: can it build a camera-and-radar satellite that survives, images, and does so again and again?

That is now the test. The next satellite has to last. The one after has to deliver data customers pay for. The constellation after that has to deliver it daily. If it works, the payoff is not simply seeing through clouds. It is turning a monsoon that blinds most satellites into something India can still watch and act on.