TARS' Embroidery Robot Cracks Automation's 'Goldbach Conjecture'

In the world of automation, some problems are so notoriously difficult they get named after unsolved mathematical proofs. But robotics startup TARS has boldly declared victory over one. At its “Needle Kung Fu” tech debut, the company pulled back the curtain on what it’s touting as the world’s first autonomous embroidery robot, which, more importantly, can transfer that fiddly finesse to the grimy realm of industrial wire harness assembly. This effectively gives the boot to a bottleneck so stubbornly persistent it’d earned the rather grand nickname, ’the Goldbach Conjecture of the industry'.

Video thumbnail

According to Founder & CEO Dr. Yilun Chen, the company’s blistering pace – considering it only popped up in February 2025 – is apparently down to a rather snazzy full-stack “DATA – AI – PHYSICS” trinity. Kicking things off is SenseHub, a wearable system that hoovers up multi-modal data from us mere mortals to give the AI its lessons. All that juicy data then gets shovelled into TARS AWE 2.0, a foundation model cooked up for end-to-end learning, allowing skills to be brilliantly generalised across a plethora of different tasks. And finally, the company’s rather aptly named “Born for AI” T-Series and A-Series robots are engineered with a laser focus to shrink that pesky chasm between pristine digital simulation and the gloriously messy reality of the shop floor.

Why is this important?

While a robot expertly stitching a logo might seem like a rather clever party trick, the actual meat and potatoes here is the bonafide, industrial-grade proof of concept for the Embodied AI Scaling Law. TARS is laying down a crystal-clear, utterly reproducible methodology for schooling robots in complex, fiddly tasks involving those notoriously tricky soft and squishy materials – a proper behemoth of a hurdle for automation, mind you. By tackling a proper, greasy-fingered industrial problem, rather than just another shiny lab demo, the company is brilliantly mapping out a scalable path from the factory floor right through to, eventually, our very own abodes. Now, if it could just master the dark art of darning socks, then, and only then, will the revolution truly be complete.