Robotics startup Tacta has stepped out of stealth, showing off a robotic hand paired with a wearable glove, according to The Information, which first reported the reveal. The pairing points to where a lot of the smartest robotics money is heading right now: giving machines a sense of touch, not just a set of cameras.
What stands out here is the combination. A dexterous robotic hand is hard enough. Pairing it with a glove is the tell that Tacta is chasing the harder problem underneath humanoid hype: how do you teach a machine to grip, twist, and handle real objects without crushing or dropping them?
🤖 What Tacta showed
- A robotic hand. The centerpiece is a hand built for dexterity, the kind of fine manipulation that trips up most of today’s robots.
- A wearable glove. The glove is the human side of the loop. A person wears it, performs a task, and the system captures that motion and touch data to train the hand.
- A touch-first approach. The whole setup leans on tactile feedback rather than relying on vision alone, which is the direction The Information’s reporting highlights.
✋ Why touch is the hard part
Most robots “see” the world through cameras and depth sensors. That works for navigation and for spotting objects. It falls apart the moment a robot has to actually handle something delicate, like folding laundry, plugging in a cable, or picking up an egg.
Humans do this with touch. You don’t look at your fingers to feel whether a grip is slipping. Tactile sensing tries to give robots that same instinct, and it’s one of the biggest unsolved bottlenecks standing between today’s lab demos and robots that are genuinely useful in homes and factories.
📊 How this fits the bigger picture
Tacta is entering a crowded, well-funded race. Humanoid and general-purpose robotics has pulled in enormous investment, with much of the attention going to full-body machines. Tacta’s focus on the hand and the data glove is a bet that manipulation, not locomotion, is where the real value sits. A robot that can walk but can’t reliably grab things isn’t much of a worker.
The glove-based training method also matters. Collecting real human demonstration data is one of the cleaner ways to teach robots complex physical skills, and it sidesteps some of the limits of pure simulation.
⚠️ Worth keeping in mind
The Information’s report is an early look, so a lot of the practical questions are still open. There’s no clear picture yet of when this reaches customers, what it costs, or how it stacks up against rivals in head-to-head tasks. Hardware demos are one thing. Reliability across thousands of messy, real-world grips is another, and that’s the bar every touch-focused robotics company still has to clear.
This is significant because it signals where the field is maturing. The conversation is shifting from robots that look impressive to robots that can actually do delicate work with their hands. Watch whether Tacta can turn a slick reveal into hardware that holds up outside the demo room. More details are available in the original report from The Information.