Small helical robots driven by a rotating external magnet offer a promising route to wireless clot retrieval in vessels that a catheter cannot reliably reach. However, two coupled failure modes limit their clinical potential. First, the rotating magnetic field is not perfectly uniform. This causes the robot to lose synchrony with the field sooner than classical theory predicts, capping the flow rates it can swim against. Second, because helical propulsion is purely axial, the robot cannot generate thrust to oppose lateral flow at branch points. This lateral drift is resisted only by the weak magnetic holding force, producing trajectory (vector-sum) errors that can send the robot into the wrong branch.
To address this, this project uses a validated, differentiable two-scale simulation, combining a confined boundary-element drag model with a finite-volume vessel-flow solver, to map where branch-selection manoeuvres under counter-flow are feasible. The ultimate goal is threefold: to define a partitioned feasibility envelope separating controllable trajectory errors from hard physical limits; to develop a trajectory-shaping control law that widens this controllable region; and to establish a definitive boundary showing exactly where smarter control is feasible, and where only stronger actuation can help.
