Stringman is one of the few robots under $1000 that reach a whole room and lift 2kg. And it's the only cable robot that can run physical intelligence models using @LeRobotHF Use it to ferry items to other robot stations, or just bring yourself a drink. A kid can drive it around with a gamepad like a drone. Pre-order at neufangled.com/stringman. We're ready to ship as soon as @Raspberry_Pi Zero 2Ws are available. There are currently 8 of them out in the wild and we're building a dataset to serve as the foundation of cleaning tasks for cable robots.

After all these years, why are people still using popups on their websites. You know if you do that, you instantly burn trust and some fraction of your users think "oh fuck you" and close the page

Truly, the limiting factor in home robotics is the number of April tags you're willing to print and put on the floor of your room.

American company: Spend year looking for PMF, pivoting, raising money, marketing, testing, prototyping. Chinese company. pick a thing that's already popular, build 100k of them, list it for sale on every market under the sun, become best selling product. How do they do this? "Pick anything and go all in" is not usually a successful strategy

Today is Saturday. I'm cleaning my house (with robot help) and reading the whole codebase to find and delete claudisms that shouldn't be there. Also how do you like the anchor redesign?

All the stuff you thought was cool in middle school is actually cool.

For some stupid reason I feel the need to respond to my critics: * It looks like it will get caught in the ceiling fan - No, you just mount it lower than the fan. Most of my videos take place in a room with a fan, it's never been a problem. * A thousand bucks!? It’s too expensive - Even a basic robot arm is twice that. Any bot that could reach your entire floor is 10k or more. Stringman is relatively cheap * It's ugly, you need to work on your image - I have no idea what image you would find appealing, My personal taste has no bearing on yours. print a different faceplate. * It looks dangerous, like its going to cut my head off and strangle me - well it wont. I can prove it, but fear ignores proof, so I can't help you * I don't want holes in my wall, can I use command strips? - no. you're better off just sqeezing your eyes shut while someone else installs it and then opening them when the holes are covered up. * Looks neat but I can't see this being useful without a remote control and video interface - you know you're right. it probably needs electricity too, I'll get on that. * You're just going to spoil your kids and teach them they never have to pick up after themselves! - this is very long term and nebulous speculation. Parenting seldom leads to the results one expects. * I call bullshit, there is no way this can work, he's just holding it like a puppet. - wow man, thanks. I'm honored that you think moving a gripper on strings is so impressive that it's beyond human engineering.

A home is a place where you are free to change anything to your liking.

Please check out Stringman on youtube youtube.com/@overengineer769…

Claude Opus 5.5 tries to "sort legos into neat rows" with Stringman and an MCP server. Doesn't get far before he's interrupted by his 3yo manager. @aihacs Much learned though! if you want to drive a cable robot with your agent, Buy one! neufangled.com/store

Watching an autonomous cdpr pick up things in my house is surreal. I dreamed that I could do this a few years ago and now here it is.

You know what would make shipping products 10x easier? If the scissors and tape dispenser were chained to the table.

A lot of useful knowledge of the object cannot be learned until it is touched. was it heavier than expected? was it tied down somehow? Maybe it was just a stain on the floor or a printed image of a toy. Sometimes the object compresses when grasped. sometimes it's sticky and has to be shaken off the fingers. What I'm saying is you should buy a Stringman and solve this problem neufangled.com/stringman

Labelling synthetic frames generated by a fine-tuned diffusion model improved my target finding. It gives the model a few more chances to learn to ignore rugs and items connected with wires, as well as ignoring things which already look like they're inside a shelf or stacked against the wall.

If my robot actually finds itself in this situation, I hope it actually helps someone. they better put a trash can in there

This is magic