Here is a genuinely different answer to a familiar problem. Inspecting a satellite in orbit — checking for damage, verifying a deployment, diagnosing a fault — usually means flying a second spacecraft up close, with all the propellant, precision, and collision risk that free-flying proximity operations demand. USC's grant proposes skipping the flying entirely: put the robot on the spacecraft and let it crawl.
The grant is US12576543B2, "Soft robotics, autonomous, space inspection, crawling robot" (issued March 17, 2026), assigned to the University of Southern California. Its CPC tags are pure robotics: B25J 15/0233, B25J 15/008, B25J 15/10 (grippers and end effectors), B25J 9/1612 and 9/1664 (manipulator control), with B64G 4/00 (space exploration equipment). "Soft robotics" is the key phrase — it means a robot built from compliant, flexible materials rather than rigid links and motors.
Why soft, and why crawl? Two reasons. First, a soft, compliant body conforms to the irregular, delicate surfaces of a spacecraft — solar arrays, thermal blankets, antennas — without the risk a rigid gripper poses of punching through or snapping something. Second, crawling along the structure means the robot is always attached: no propellant to maneuver, no relative-velocity error that could turn an inspection into a collision. It trades the freedom of free flight for the safety and fuel-economy of staying put.
This is the dependent-claim-is-the-moat situation in a new domain. The broad idea — inspect by crawling — is intuitive; the value is in the specific grippers and compliant mechanisms that let a robot hold onto and traverse a spacecraft in microgravity without damaging it or itself drifting off. Those are exactly the B25J gripper claims on the grant.
The scope and maturity note matters here. This is university research — an early-stage concept protected as IP, not a flight-qualified servicer. Soft robotics in the thermal extremes and vacuum of space is a hard materials problem, and a crawling inspector solves only the tasks where being attached is an advantage; it cannot service a satellite it can't first reach. The patent stakes out an approach; it does not claim a demonstrated mission.
For the servicing lane, the interest is the diversity of approaches now being patented. Free-flying multi-agent systems, single exquisite servicers, and now surface-crawling soft robots are all staking claims to in-orbit inspection and maintenance. That breadth is itself a signal: the market believes servicing is coming, and the IP land grab — across radically different mechanisms — is well underway.