The usual way to make a soft robot is to build the machine first and then cover it: actuators, sensors, and tubing assembled into a device, with fabric added around the outside. Rice University's Vanessa Sanchez is inverting that. In her lab, the textile is the robot — sensing, movement, and mechanical function encoded directly into the knit during manufacturing.

Programming a Stitch

Sanchez directs Rice's texlab, which designs robotic textiles and soft wearable systems. The core idea is that a knitted structure is already a programmable system: by controlling yarn selection, stitch pattern, and three-dimensional knit architecture, you determine how the resulting fabric behaves mechanically.

Her textiles bend, twist, contract, or inflate — and sense their own movement and the user's interaction with them — because functional fibers, pneumatic channels, and embedded sensors are built into a single knitted structure rather than bolted on afterward. The payoff is the elimination of bulky external components. A conventional soft robot needs a compressor, valves, wiring, and rigid mounting points. A programmable textile can carry much of that function in its own geometry.

It is also, notably, a manufacturing story. Industrial knitting machines are mature, fast, and cheap. If robotic function can be expressed as a knit pattern, then producing soft robots starts to look less like precision assembly and more like running a textile mill — a very different cost curve.

The Toyota Fellowship

Sanchez received the Toyota Programmable System Innovation Fellowship for a project titled "Programmable 3D-Knitted Soft Robotic Textiles for Human-Centered Mobility." The 12-month effort integrates sensing and actuation within knitted structures, with prototypes evaluated for performance, manufacturability, and safe human interaction in collaboration with Toyota researchers.

That an automaker is funding this is a useful signal about where it might land first. The envisioned applications cluster around the human body and the vehicle interior: car seats that conform to their occupant, adaptive seating surfaces, assistive clothing that supports movement, wearable localized assistance, and soft safety systems that deploy to absorb impact. An airbag is, after all, already a textile that actuates — just a crude, single-use one.

An Unusual Path

Sanchez's background explains the approach. She studied fashion design at FIT before moving into fiber science at Cornell, then took advanced degrees at Harvard and Stanford. Very few roboticists arrive by way of garment construction, and it shows in the framing: she treats knit architecture as the design language rather than as packaging for hardware built elsewhere.

Why It Matters Here

Houston's robotics reputation is built on heavy, hard machines — subsea vehicles, drilling automation, humanoids for shipyards. Soft robotics is the opposite end of the field, and Rice is where most of the region's work on it lives. The two connect more than they appear to: the hardest unsolved problems in industrial robotics are about safe, compliant contact with people and irregular objects, which is exactly what rigid machines are bad at.

There is also a neat inversion worth noting. We covered Persona AI's partnership with Under Armour, where performance fabrics are used to protect an industrial humanoid from a shipyard environment. At texlab the relationship flips: the fabric isn't armor for the robot, it's the robot itself. Two Houston-area projects, opposite ends of the same question about what textiles and machines can be to each other.

Explore more of the region's robotics research and companies in our directory and community hub.

Based on Rice University news coverage of the texlab and the Toyota Programmable System Innovation Fellowship, September 2026.