The Rocket Is the Moat: NordSpace Bets Its VLEO Future on a Launcher That Hasnt Flew Yet

NordSpace has never launched anything to orbit. It plans to replace its satellites every three years with a rocket that has never flown.

The Canadian startup — which also builds rockets and runs a spaceport, because apparently one unproven vehicle is not enough — announced April 21 a $183,000 DND contract to fund early work on Kestrel, a very low Earth orbit imaging constellation designed to deliver 10-centimeter resolution from roughly 300 kilometers up. That is spy-satellite-grade Earth observation from a commercially operated constellation. It is also, by NordSpace's own technical documentation, scheduled to burn up in the atmosphere in roughly three years.

The replacement vehicle is called Tundra. Tundra has never flown. Its first flight is slated for 2028, according to the company. (SpaceNews)

This is the dependency problem. VLEO altitudes produce intense atmospheric drag — satellites must constantly propel themselves forward to stay in orbit, like pedaling a bicycle into a headwind. NordSpace acknowledges this means frequent replenishment and a replacement cadence conventional LEO operators do not face. At standard LEO altitudes, satellites last about five years. At VLEO, three is the design target. (SpaceNews, The Conversation)

NordSpace's answer is vertical integration. The company is building the satellite bus, the imaging payload, the edge processing stack, the thruster, the orbital launch vehicle, and the spaceport. Founder and CEO Rahul Goel has invested roughly CA$5 million of his own capital into the company, which he founded in 2022 in Markham, Ontario. NordSpace has also taken stakes in domestic space companies through a venture arm, including the Earth observation operator Wyvern. The full stack is the argument: VLEO economics only work if you own the replenishment chain. (SpaceNews, Wikipedia)

That argument is coherent. It is also unproven.

NordSpace's suborbital test vehicle, Taiga, attempted to launch twice from the Atlantic Spaceport Complex in August 2025. Both attempts scrubbed — one due to a ground safety system fault, one due to an automatic misfire detection system that later proved to be a false trigger. The orbital vehicle, Tundra, is scheduled for 2028. Until then, NordSpace is dependent on SpaceX rideshare for any orbital deployment. The company plans to launch Terra Nova, a LEO imaging pathfinder, this fall via SpaceX rideshare. Terra Nova carries a proprietary NVIDIA-powered imaging system called Chronos, with edge-AI processing that NordSpace says reduces downlink data volume by a factor of 100, and a thruster called Zephyr for station-keeping tests. That pathfinder will validate technologies earmarked for Kestrel. (SpaceNews)

NordSpace has received more than 10 million Canadian dollars (about $7.3 million USD) in grants so far, mostly through Launch the North — a Canadian defence initiative to build sovereign launch capability with a $105 million IDEaS challenge program. The $183,000 DND contract is the most recent commitment and the one announced April 21. The larger figure is already committed. (SpaceNews, Government of Canada)

The US company Albedo launched its first VLEO imaging satellite in March 2025 and already holds a National Reconnaissance Office contract for 10-centimeter resolution imagery. NordSpace says no commercial product currently offers this resolution — a claim that is technically defensible if you exclude government-contracted systems from the definition of "commercial," which NordSpace appears to do. Albedo operates in the same orbit class, faces the same replenishment challenge, and does not own a rocket company. (Wikipedia)

NordSpace will get its first real data on whether the full-stack VLEO model works when Terra Nova reaches orbit this fall. The Kestrel deployment timeline and the vertical integration thesis both depend on what that pathfinder shows and whether Tundra stays on schedule for a 2028 first flight. The rocket is not part of the story. It is the story.