X-Hinges prints sensing into flexible objects, from robot-teaching gloves to lamps

MIT and Tianjin University researchers use different conductive materials to separate motion sensing from internal wiring.

Researchers at MIT CSAIL and Tianjin University have introduced X-Hinges, software for designing flexible three-dimensional printed objects with sensing embedded in their joints. MIT published details on 22 October, ahead of the work’s planned presentation at the ACM Symposium on User Interface Software and Technology.

The system uses separate materials for sensing elements and wire-like traces, reducing interference between their signals. Users can select shapes, connect parts and choose movement along as many as three axes before fabrication.

Demonstrations include a glove that sends finger-motion information to a robotic hand, a lamp that changes brightness when squeezed, and a device that recognises objects from tactile measurements.

Designing movement and measurement together

The interesting shift is the joint treatment of shape and sensing. In many prototypes, a designer first makes a physical mechanism and then finds places to attach sensors. X-Hinges begins with the intended movement and places sensing where the printed object bends.

For a glove, that means the mechanical design and measurement design are closely related. The useful signal is not simply that a finger moved; it is how its joints moved and how that motion should map to another system.

The same principle applies to an interactive lamp. A deformation becomes an input, and that input changes a visible output. The object can provide a control surface through its own material form, rather than requiring a separate switch or touch panel.

A bridge between demonstration and machine action

A motion-sensing glove can capture how a person handles an object. Turning that observation into a robot action requires a mapping between two different bodies: the human hand and the robotic mechanism.

That makes the glove a possible part of a larger teaching workflow. A person demonstrates a grasp, the system records movement, and a robot uses the resulting information to attempt a corresponding action. Repeated demonstrations can expose differences in grip, object size and timing.

The value is not limited to humanoid robotics. Flexible controls could provide inputs for games, assistive devices or custom interfaces. A designer can choose a physical interaction that fits a user’s task, then arrange the measurements needed to interpret it.

Calibration remains part of fabrication

The researchers report that nominally identical prints can produce different electrical readings. Their current solution uses computer vision to recalibrate outputs, while built-in signal recalibration remains a future goal.

That detail makes calibration an ordinary stage of making the object. A successful print supplies its geometry, but its sensing behaviour still has to be characterised. The final usable artifact is therefore the mechanism together with the mapping from its signals to meaningful motion.

For a workshop or research lab, the practical appeal is a shorter path from an interface idea to a functioning physical prototype. Changes in shape, stiffness and movement can be explored within one design process, with sensing treated as part of the object.

X-Hinges also points toward a wider direction in embodied computing: devices that acquire information through the way their materials move. The immediate contribution is a concrete design-and-fabrication system, giving researchers more ways to experiment with that relationship.

Sources

- [This design software reimagines everyday objects as self-aware devices](https://news.mit.edu/2026/design-software-reimagines-everyday-objects-as-self-aware-devices-1022) — MIT News · 2026-10-22