The grant, then the economics. On June 9, 2026, the Regents of the University of Michigan were granted US12650115B2, "Hall-effect thruster system with applied counter-torque." A Hall thruster is electric propulsion: it accelerates ions to produce thrust far more efficiently than chemical rockets, at the cost of lower thrust over longer burns. The classification sits in F03H 1/0062, the ion-thruster art, with a B64G 1/413 spacecraft-power tie.

Read the segment table this implies. A satellite's economics come down to how much useful payload it carries versus how much mass it must spend on staying alive — propellant for station-keeping and orbit changes. Electric propulsion shifts that ratio. Less propellant mass means more payload, or a smaller, cheaper bus, or a longer operational life. Every one of those moves the cost-per-kilogram-per-year number that an analyst uses to judge whether a constellation can ever clear its cost of capital.

“A Hall effect thruster system includes a thruster body and a diffuser configured to apply a torque to the thruster body during operation.”— U.S. Patent No. 12,650,115 source

The "counter-torque" wrinkle in this particular grant points at a known operational headache: thrusters can induce unwanted rotation, which then costs you reaction-wheel capacity or extra control fuel to fix. Solving that at the thruster level is, in business terms, removing a tax on the rest of the spacecraft's budget.

The measured caveat this beat insists on: a university patent is not a fielded product, and the path from a Regents grant to a flight-qualified thruster in a commercial constellation is long and uncertain. We are reading the direction, not declaring a winner.

But the direction is the story. The reason propulsion IP keeps clustering — across universities, labs, and primes — is that it is one of the few levers that moves satellite unit economics structurally rather than incrementally. When the cost-per-kilogram math improves, marginal constellation business plans become fundable. The thruster patent and the financing decision are, in the end, the same conversation about whether the numbers close.