Concentric tube robots are surgical instruments made of nested, pre-curved tubes that can be telescoped and rotated relative to each other. By sliding tubes in and out, the robot's tip traces a path through tissue. The tubes are typically nitinol — superelastic, capable of holding complex curvatures while remaining flexible enough to navigate anatomy. The problem: you can control which tubes are extended and how much they're rotated, but you can't directly observe the resulting shape. The tubes interact mechanically — their curvatures add, subtract, and couple in ways that depend on their relative positions and the forces they encounter.
Note (2026-11-14; read against the paper's full text): the essay says one fiber gives complete three-dimensional shape information. The paper reconstructs a two-dimensional shape: it attaches a single OFDR fiber to a flat NiTi wire (the shape sensing assembly), calibrates strain against curvature in a planar jig, and reconstructs the 2D shape; its own earlier reference is titled as being for planar bending. The application is steerable drilling for vertebral compression fractures, with the assembly housed inside the internal channel of a flexible drilling instrument that passes through the pre-shaped NiTi tube, tested in free bending and in drilling J-shaped trajectories in synthetic bone phantoms (maximum average tip errors of 1.32 mm in free bending and 1.73 mm in drilling, normalized shape error under 1.5 percent in the free-bending tests). The reason the paper gives for a single fiber is simpler and cheaper fabrication and loose insertion without adhesion to the instrument, compared with earlier assemblies that needed several fibers in machined channels; I found no statement that the narrow lumen forces the single-fiber design.
Model-based shape prediction exists but accumulates error when the robot contacts tissue, when tubes interact in ways the model simplifies, or when manufacturing tolerances shift the pre-curvatures from their nominal values. What you need is a direct measurement of the shape.
A single optical fiber threaded through the concentric tube assembly provides this. Optical Frequency Domain Reflectometry (OFDR) measures strain along the fiber at thousands of points. Strain maps to curvature. Curvature along the fiber maps to the three-dimensional shape of the robot. One fiber, one measurement channel, complete shape information.
The single-fiber approach is notable because concentric tube robots have very narrow lumens — there isn't room for multiple sensing elements. One fiber fits. Multiple fibers might not. The constraint of the device geometry forces the sensing architecture, and the OFDR technique is one of the few that can extract distributed three-dimensional shape information from a single fiber. The measurement matches the constraint.