Soft grippers solve one problem and create another. TRI's grant US11389968B2 ("Systems and methods for determining pose of objects held by flexible end effectors," issued July 19, 2022; inventors Naveen Suresh Kuppuswamy, Alexander Alspach) addresses the created one: when a flexible gripper grasps an object, the compliance that makes the grasp gentle also makes the object's exact position uncertain. The claim estimates that pose anyway.
The mechanism uses tactile and force sensing to infer where the held object is despite the gripper's deformation. B25J 15/12 marks the soft/fluid gripper; B25J 13/082 and 13/084 mark the touch and force sensors; G01L 5/228 specifically covers measuring forces in a gripping action. Together they describe a system that reasons from contact forces back to object geometry — the inverse problem a rigid gripper never has to solve.
For the manipulation beat, this is a beautiful example of how an enabling technology spawns a derivative IP need. Soft grippers (Toyota, Soft Robotics, and others fenced the hardware) are great for delicate handling, but they break the assumption that gripper pose equals object pose. A claim that re-establishes object pose for compliant grasps is the missing piece that makes soft manipulation usable in precision tasks.
From a portfolio angle, TRI's manipulation filings consistently target the hard, specific sub-problems of human-environment robotics — the compliant, uncertain, real-world cases. A pose-from-compliant-grasp claim is exactly that flavor: narrow, deep, and tied to a capability (soft, safe manipulation) that humanoid and home-robot builders all want. It's research IP with a clear product trajectory.
Caveats. Tactile pose estimation has academic prior art, and the grant turns on the specific inference method in claim 1, not on the idea of sensing pose through a soft gripper. The G01L and B25J 15/12 codes mark the soft-sensing domain; they don't define the algorithm. Read the independent claim for how forces are mapped to pose.
For the file: a soft-manipulation sensing grant that closes the pose-uncertainty gap compliant grippers create. Pull US11389968B2 on PatentBear, read claim 1 for the force-to-pose inference step, and file it as enabling IP for precision soft manipulation.