← All Projects
In Preparation · Science Robotics
Procedural Design Whole-Body Tactile Sensing

GenTact Toolbox

Robot Skin for All Robots

A computational pipeline for generating custom, context-driven whole-body tactile skins that conform to any robot geometry.

Abstract

Robots lack skin, and existing sensor attachment is tedious, robot-specific, and hard to reproduce. We propose multi-material additive manufacturing instead: one print yields a complete skin unit — embedded sensors, conductive traces, compliant structures — straight from a digital design file.

Authors

Caleb Escobedo*†, Carson Kohlbrenner, Anna Soukhovei, Klara Nitsche, Artemis Shaw, Nikolaus Correll, Alessandro Roncone
Department of Computer Science, University of Colorado Boulder  ·  *Corresponding author  ·  Equal contribution

Six Sensing Modalities

One print can embed fundamentally different sensor physics in the same skin unit, on a shared microcontroller.

Optical Time-of-Flight

8×8 multizone proximity up to 4 m, direct from the robot surface.

Self-Capacitive

Conductive PLA electrodes printed into the dermis — senses contact off the print bed.

Mutual-Capacitive

Electrode pairs for higher-resolution contact localisation.

Electrical Impedance Tomography

Distributed contact maps over a continuous conductive medium.

Magnetic

Hall-effect sensors in a flexible magnet lattice for force sensing on curved surfaces.

Hybrid ToF + Capacitive

Both modalities in one skin unit on a shared microcontroller.

Demo — real-time capacitive touch response on a GenTact prototype

Three Evaluation Axes

Repeatability

Consistent Across Prints

Self-capacitive skins achieved 139 ± 19 ADC peak response with 8.6% inter-prototype CV across three identical MMAM prints.

Scalability

Any Robot, Any Size

Full-body skins fabricated for a robot arm, quadruped, and humanoid — from small sensor patches to complete body coverage.

Compatibility

Multiple Modalities, One Skin

Different sensing physics co-exist in a single skin unit on a shared microcontroller — calibration handled in the digital design.

Isaac Sim — real-time contact simulation and automatic sensor placement refinement for specific tasks

The Fabrication Pipeline

  • Design — sensor layouts generated in the open-source GenTact Blender addon, before anything is manufactured.
  • Simulate — 1:1 import into Isaac Sim to refine placement against specific manipulation or HRI tasks.
  • Print — a Prusa XL lays down rigid, soft, and conductive materials in one uninterrupted job.
Materials and tooling, in detail

Five independent nozzles extrude rigid PLA (dermis), soft TPU (epidermis), conductive PLA (electrodes and traces), and support material in one uninterrupted print. Geometry Nodes place the sensors and route the internal wiring.

Blender Addon — procedural tactile skin generation using Geometry Nodes, conforming sensor layouts to any robot mesh

Future Applications

The pipeline generalises wherever compliant, geometry-aware sensing is the bottleneck.

Prosthetics & Bionic Limbs

Skins at the prosthetic socket interface, coupling contact signals to nerve stimulation to restore the felt sense of touch.

Proximity-Aware Manipulation

ToF skins let robots react before contact — pre-grasp shaping, whole-arm collision avoidance, safer HRI.

Clinical & Assistive Robotics

Bedside robots that detect unintended contact and respond compliantly during clinical procedures.

Soft & Wearable Robotics

Exoskeletons and supernumerary limbs, where the sensing layer must conform and survive repeated flex cycles.

Full-body procedural tactile skin deployed in a human-robot interaction scenario on the Franka Research 3