Hacker News new | past | comments | ask | show | jobs | submit | Wolley's comments login

I've been working on a modular vacuum cleaner robot for around 5 years:

https://wolley.se

I have done it in my spare time but the dream is to be able to sell it as a kit and work on it full time.


Side note:

People here often say that hardware is hard, but I think OSHW has the upside of de-risking it too (although I might be somewhat biased in my enthusiastic support of OSHW). If other people are producing it (from 3d printing at home to commercial manufacturing), you don't have manufacturing risks at all. The large costs and difficulty involved are probably mostly around manufacturing. Also by extending the development phase and working closely with customers you can find a product-market fit more naturally (and iterate if it fails... or just abandon it, I guess). If all goes wrong you've contributed an open source project...

The hard part of course in that case is convincing your users and any company that decides to manufacture your hardware to pay you. But not unheard of, see LumenPNP (https://opulo.io/products/lumenpnp)

My dream is a completely (or almost completely) open society, where most development is open and people collaborate more or less freely, and we have good mechanisms (and enough good sense individually?) to reward and invest in all this work.


Awesome. I had a similar idea. Maybe we could join forces one day?...

I thought about hacking and modding existing robots (I have one at home that has great hardware, but the software is terrible!).

Definitely set up a Liberapay account to get support (if you need economic support), although I guess most people would want to see proof of concept first.

One thing that would be great is to crowdsource the software, and maybe the training data for ML. The open source community got together and made the top chess and go engines with community GPU training, why not do the same for robot vacuums?

Good luck! :)


Thanks! Sure, there's a lot of more work to do, especially the software for the "brain". I have mostly focused on the mechanics and electronics since that's the most exciting/new for me.

I will in a couple of months have a skeleton/chassi finished and then the "only" thing left is to write the software to make it autonomous. The idea is to keep it really simple in the beginning and let it run on a Raspberry Pi Zero W. Using a Raspberry Pi opens up the platform for more people and software.

Yes, I would want to have something that works fine before starting to sell components, parts or accepting donations etc.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: