GelSight raises $10M Series B for its imaging system that uses an elastomeric-based sensor and computer vision to produce highly accurate 3D maps of surfaces
Jeremy Bogaisky / Forbes :
Context & Ripple Effects
GelSight's $10M Series B lands in the middle of a sustained venture run on machine perception: Veo Robotics' $12M Series A in 2017 and EyeSight's driver-monitoring round in 2018 preceded it, and UltraSense Systems raised a $20M Series B just weeks later for ultrasound sensing through metal, glass, and wood. Investors are funding an entire sensor layer around the camera.
The round also reads differently in hindsight: the supplied relationship shows Meta later working with Wonik and GelSight to develop and commercialize tactile sensors for AI, including the Allegro robotic hand and Digit 360 tactile fingertips — meaning this 2020 raise backed technology that became a component in frontier robotics research.
First-order effects
- The $10M gives GelSight capital to move its elastomeric-sensor-plus-computer-vision imaging from niche instrumentation toward volume manufacturing of 3D surface-mapping sensors.
- Meta's work with Wonik and GelSight on the Allegro hand and Digit 360 fingertips gives the startup a marquee design-in customer, anchoring revenue in robotics rather than one-off research sales.
Second-order effects
- UltraSense's concurrent raise shows competitors attacking adjacent sensing problems — gesture through solid materials versus GelSight's fine surface topology — forcing buyers to assemble multi-modal perception stacks from separate vendors.
- Robotics and device makers gain an off-the-shelf touch-perception option, shifting integration effort from building custom sensors to choosing among them.
Third-order effects
- If tactile sensors follow the path cameras took — standardized components bought rather than built — robot platforms will consolidate around common perception hardware, with startups like GelSight competing on sensor physics while the software layer absorbs the differentiation.
- The pattern of perception-specific rounds (vision enhancement, ultrasound, gaze tracking, touch) points toward perception infrastructure becoming its own funded category, distinct from the AI models that consume its output.
The trend: Machine perception is expanding beyond cameras into touch and through-material sensing, with successive venture rounds funding a dedicated hardware sensor layer for robotics and devices.