Even with breakthroughs in language and vision AI, the physical world remains frustratingly out of reach for machines. Even the most advanced large language models can write code, compose poetry, or simulate conversation, but they still struggle to open a door, sort through clutter, or fold a towel. This chasm between cognitive capability and physical dexterity is one of the most formidable barriers in robotics. Tacta Systems, a Palo Alto-based startup, is betting that the solution lies in what it calls “Dextrous Intelligence,” and it just raised $75 million to prove it can work.
The company’s Series A round, led by America’s Frontier Fund and SBVA, with participation from Matter Venture Partners, B Capital, EDBI, and others, builds on an earlier, undisclosed $11 million seed round. The total funding underscores growing investor appetite for startups focused on bridging the gap between AI and real-world physical manipulation. Tacta’s pitch is the idea that robots need not just vision and logic, but also a kind of embodied intuition: an ability to perceive, react to, and operate within the physical world with the fluidity and finesse of human hands.
A Nervous System for Machines
Tacta’s approach combines software, hardware, and AI into what it calls a “smart nervous system” for robots. Its proprietary Dextrous Intelligence platform enables machines to sense tactile information, adapt to varying conditions, and perform complex tasks with speed and flexibility. “AI models have become incredibly sophisticated in working with text and video, but much of the physical world remains incomprehensible to them,” said Andreas Bibl, Tacta’s co-founder and CEO, in a press release announcing the funding.’
Tacta’s technology aims to solve a fundamental limitation in today’s robotics: the lack of sensory intelligence that allows for subtle, context-aware movement. Industrial robots, for instance, can execute repetitive motions at high speed, but they often require highly structured environments. Even slight deviations in object placement or shape can cause failures.
This capability is especially important as the robotics industry shifts from closed, factory-line automation to more dynamic, unstructured environments like warehouses, hospitals, or even homes. The company believes that dextrous manipulation: the ability to handle a variety of objects and tasks without explicit programming, is essential for the next wave of general-purpose robots.
The Physical Gap in AI’s Growth
Tacta’s rise comes at a moment of intense focus on what researchers call “world models”: AI systems that simulate and reason about the real world, often by integrating sensory data from images, videos, text, and audio. These models, pioneered by companies like DeepMind and Meta, promise to bring a deeper form of understanding to AI agents. But so far, most applications of world models remain virtual: generating more realistic video, forecasting future events, or simulating physics in video games.
What’s missing is physical grounding. As Snap’s former AI chief Alex Mashrabov put it, even advanced video models can get tripped up when the world behaves in unexpected ways. A feather that falls like an anvil or a ball that floats instead of bouncing is a sign that the model lacks a true grasp of physics. Robots, unlike videos, must act on those physics in real time, with stakes ranging from factory uptime to human safety.
This is where Tacta’s strategy intersects with broader AI development. By focusing on tactile perception and real-world interaction, the company positions itself as a crucial layer between high-level intelligence and mechanical execution. In other words, it’s not building the brain, but the sensory-motor system that allows the brain to interact with the world.
That framing aligns with where many in the industry believe robotics is headed. Nvidia CEO Jensen Huang recently described robotics, alongside AI, as one of the company’s two largest long-term growth opportunities. “We’re working towards a day where there will be billions of robots, hundreds of millions of autonomous vehicles, and hundreds of thousands of robotic factories,” Huang said at Nvidia’s 2025 shareholder meeting. Tactile intelligence, as envisioned by Tacta, could be a prerequisite for those robots to function beyond the lab or the line.
Tacta’s leadership team includes veteran entrepreneurs with experience building and scaling hard tech ventures, which may prove critical in navigating the notoriously slow and capital-intensive world of robotics. Unlike pure software startups, Tacta must design and integrate specialized sensors, test in physical environments, and collaborate with manufacturers. The ability to execute across disciplines is often what separates promising robotics startups from those that never ship.
So far, the company has kept details about specific product applications under wraps. But the language in its funding announcement suggests a focus on industrial use cases: factory work, logistics, and environments where human dexterity is currently indispensable. That mirrors the playbook of companies like Apptronik, whose new subsidiary Elevate Robotics is developing non-humanoid systems for heavy-duty tasks.
The potential impact is not lost on investors. Will Fung of Woven Capital, which also joined the round, noted that the lack of an “internet-scale dataset for touch and manipulation” remains a fundamental gap in robotics. “By advancing both hardware and software, Tacta is laying the foundation to bridge that divide,” Fung wrote in a LinkedIn post.
