Work for a Member organization and need a Member Portal account? Register here with your official email address.
What if clothing and textiles emerged in situ—woven by robots around the body itself, shaped by anatomy, motion, and material behavior rather than predefined geometry?
InSituWear is a form-finding, on-body robotic fabrication method that melt-draws low-temperature thermoplastic microfilaments directly onto the human body, eliminating the slow, multi-stage processes of scanning, modeling, and assembly while enabling custom-fit, waste-free fabrication with embedded textile intelligence.
The InSituWear method is structured around a bespoke computational pipeline that translates design intent into safe, executable robotic motion—defining coarse body geometry, safety boundaries, toolpaths, clash detection, and synchronized fabrication in a single, human-facing workflow.
Complementing this pipeline, a modular end-effector suite enables robotic melt-drawing through interchangeable single- and multi-needle tools with localized heating and compliance, supporting both fine-grained control and scalable textile fabrication directly on the body.
To quantify strand quality and controllability, we performed robotic pull-out tests that draw single filaments from molten material under fixed dip depth and pull height. The study isolates how fabrication dynamics shape fiber formation before full textile patterning.
We characterize how robotic parameters, especially pulling speed and material composition, directly control filament diameter, producing fibers from hair-thin strands to thicker structural filaments. This establishes a tunable mapping from computation to material outcome for performance-driven textiles.
A pattern-making setup translates toolpaths into wrapped textiles by varying strand orientation, spacing, and looping logic around cylindrical targets. We prototype multiple wrapping families—parallel, crossing, and diagrid—revealing how geometry governs cohesion and mechanical behavior.
Tensile testing demonstrates that melt-drawn InSituWear textiles perform within the mechanical range of common commercial fabrics, while offering tunable behavior through computational patterning.
Beyond structure, fibers can embed responsive behaviors through additives and integrated conductive elements, enabling textiles that react to heat and touch. This positions intelligence as a native material property, formed during the fabrication process.
Cross-Pattern Sleeve: The sleeve prototype demonstrates safe on-body fabrication at full wearable scale, executing controlled woven patterns directly around the arm. It validates stability, comfort, and repeatability while maintaining precise pattern intent.
Thermochromic Glove: The glove prototype integrates textile intelligence through thermochromic PCL and embedded nichrome wires, producing localized activation triggered by hand gestures. It demonstrates responsive behavior as an intrinsic part of the fabricated fiber network.
Furniture & Architecture: The method extends beyond the body by wrapping open 3D frames, producing lightweight structural textiles that can be reinforced for strength. These explorations suggest a pathway from wearable-scale fabrication to fiber-based spatial systems in furniture and architecture.
Sergio Mutis (Research Lead & Robot Control), Justin Wan (Computational Design & Robot Control), Berfin Ataman (Mechanical Evaluation), Abigail Suk (Fiber Formation & Material Intelligence), Ayah Mahmoud (Fiber Formation Studies), Jiaji Li (Technical & Research Overview), Frank Haotian Cong (Early System Development), Paolo Salvagione (Mechanical & Technical Engineering Consultant)
Behnaz Farahi (Critical Matter Group Director)
This work is partially supported by the MIT–LUMA Grant.