Haptic interfaces / physical AI

The Last Millimeter Problem.

You can’t have physical AI without touch.

The highest-resolution haptic feedback actuators in the world, for robot teleoperation, contact-rich demonstration data, and the people who collect it.

01 The stakes

The most valuable work in the world is done by hand.

$50T

of global economic activity flows through human hands every year

$15T

flows through navigation: the trillions chased by autonomy

Manipulation is a 3× bigger market than mobility, funded at a fraction of it. Source: Cortical Ventures

02 The shift

Robots are taught, not programmed.

Three years ago, you programmed waypoints. Today, robots learn by watching people work. Every major AI lab has converged on teaching by human example.

But there’s a problem.

03 The problem

Teaching breaks at the last millimeter.

A teleoperator without touch demonstrates numb, like plugging in a USB‑C cable wearing oven mitts. Demonstrations without touch produce data without touch.

Teleoperation without haptic feedback is…

Slow.

Unnatural.

Unsuccessful.

04 The argument

Vision is not enough.

You can move a robot the first 999 millimeters with vision. The last millimeter — plugging in a USB‑C cable, threading a screw, feeling a part click home — is the domain of touch.

Feel what robots feel.

05 The demo

The highest-resolution haptic feedback actuators in the world.

A gloved operator, a robot hand, and an actuator array rendering every contact, live, side by side.

Contact us for special projects Our actuators make new things possible
06 The instrument

Feel the contact.
Feel the pressure.

Row of electroosmotic haptic actuators on black glass, macro photograph
6 mm · electroosmotic actuator

One hundred times smaller than pneumatics, micro-hydraulic electroosmotic actuators are silent and small enough to put variable pressure haptic feedback under every fingertip.

6 mm ⌀

Actuator diameter

15 ms

Response time

128 lvl

Pressure levels

07 The path
Someday, physical AI will earn full autonomy. Today, we need mixed autonomy.

Thread 01 / Data

Touch is the data physical AI needs.

Manipulation models fail at contact because their training data never contained it. Operators who can feel demonstrate correctly, and capture the contact dynamics no camera can see.

Thread 02 / Deployment

Autonomy has to be earned.

Tesla and Waymo climbed the autonomy ladder for a decade: driver in the loop, then supervised, now hands off but still calling back to a human command center. Manipulation is now climbing the same ladder, and haptics allow people to steer.

1:1 1:5 1:30 One operator, more robots every year. The future of robots needs command centers. We build the touch inside them.
08 Start building

Lansing Dev Kit

Electroosmotic haptic actuators you can integrate into hardware you already own. Fingertip-scale, solid-state, silent.

For robotics labs · teleoperation teams · manipulation researchers · haptic interface builders

Not yet on sale. Early access by waitlist.

Join the waitlist

Built with researchers and engineers from

Carnegie Mellon UniversityNorthwestern UniversityIntuitive Surgical
09 Choose your door

Door 01

Start building

The Lansing Dev Kit puts variable pressure haptic feedback in your hands. Everything needed to get started fast: actuators, controller, Python SDK, and docs.