Folding laundry is one of those household tasks that feels like a win when it’s done but a grind while you’re doing it. Figure’s Helix hopes to make that chore a thing of the past.
Helix’s ability to fold laundry comes from its Vision-Language-Action (VLA) model, a system that lets it see, understand and act on verbal instructions. When someone says “fold the towel”, Helix doesn’t just hear the words. It uses cameras to scan the pile, processes the request and moves its multi-fingered hands to pick up a towel, smooth it out and fold it into a neat square. The video shows Helix working through a stack of towels, adjusting its grip for each one’s shape and texture. This adaptability is key because towels aren’t rigid – they flop, twist and tangle making them a nightmare for most robots. Helix however navigates these challenges with eerie competence.
- 【BRILLIANT GIFT IDEA FOR KIDS】It's a big surprise to kids as the robot is up to 15.8 inches in height. With various pioneering ways to play, it...
- 【STEM LEARNING ROBOT】With 10 emojis, 9 flexible joints, and 10 songs, this robot moves more dynamically than typical toys—making playtime more...
- 【DIVERSE FUNCTIONS】Gymnastics, storytelling, dance, music, and recording—the Ruko robot enriches childhood with creativity and early artistic...
What makes this possible is Helix’s end-to-end neural network, a single system that handles everything from seeing the towel to moving the robot’s fingers. Unlike older robots that needed separate programs for each task, Helix uses the same setup it uses for other jobs like sorting packages or setting tables. The only change for laundry? New data fed into the system. Figure’s team trained Helix with a dataset tailored to folding, teaching it to recognize towel edges, pinch corners and even untangle messes without altering its core design. This means Helix can switch from warehouse work to household chores without needing a hardware overhaul.
A towel’s shape changes with every touch and there’s no fixed spot to grab it. Helix’s fingers can perform delicate tasks like tracing an edge or smoothing a surface. Even errors, such as picking up two towels and returning one to the pile, can be fixed by the robot. This is possible because of Helix’s ability to process visual and tactile data in real time.
Beyond folding, Helix has personality. It looks you in the eye and uses hand gestures when interacting with people, so it doesn’t feel like a machine and more like a helper. These little details, shown in the video, are part of Figure’s bigger goal: robots that blend into your home. The company’s CEO, Brett Adcock, has said that Helix’s skills are being tested in real homes, not just labs. A robot that can fold laundry today and cook dinner tomorrow is cool, and pulls you into the possibilities.
This isn’t about replacing humans but about freeing up time. Helix’s laundry folding demo is impressive, but a stepping stone. Figure wants to scale its system to do more complex tasks—think setting a table or assembling furniture—using the same platform. The robot can learn new skills with just more data so it could do anything from helping in a factory to tidying your living room.
[Source]
