The hardware is an issue, not because it's bad but because it's massively expensive to buy the components piecemeal.
You can purchase a lidar vac for £70-80 now. Even if you only replaced the brains, that's a quarter of the price of a Oomwoo. The only upgrade I'd want is self-emptying. You'd probably have to relocate the charging contacts but it seems highly achievable.
Or you could break up an existing vac for the parts. You'd get the lidar, bumper, ToF, cliff sensors, motors and wheels, perhaps even some seals for your printed parts. Again, much cheaper, especially if you shop on the used market (I can get a whole working vac for the price of new wheels). All these robots use common parts so the risk of getting it wrong is very small.
My point is perhaps they could coalesce around a common white label option unit or set of parts currently sold as a vacuum.
Gagguino is a great example of this approach, licensing scandal aside†. Espresso machines are expensive and not because the software is particularly clever. They are electrically simple, but mechanically there's a lot of pressurized plumbing that you really don't want to DIY.
The kit is a control board for the pump and boiler, and some add-on sensors for temperature and pressure. The "high end" features that it enables are almost entirely software driven, the main one being temperature control via PID. I've seen even simpler mods for other machines that bypass the "brew" button so you can do things like connect a bluetooth scale to enable brew-by-weight on a machine that doesn't support it, or add a shot timer.
The commercial version of this would be the Decent, but it costs 3x as much. I would love something like this for my robot vacuums. Valetudo is minimally invasive, but there's no reason you couldn't control the vacuum + wheels, but navigation is hard and those sensors are much more complex (can you even access the camera and undistort the image?)
> pressurized plumbing that you really don't want to DIY
They're fine to DIY if you're sensible about it. If you can build it, you can build a simple hydrostatic pressure tester. Pump, valves, tubes and a gauge. No computer parts. Pump the tank up to 1.5x rated pressure. This is industry standard.
But of course the age old saying applies: know what you're doing.
I own three robot vacuums from two different manufacturers that are nearly identical. Exact same charger station that seems to match what Oomwoo uses. My point is that Oomwoo seems to use at least some common parts already.
Yeah I don't think fully open sourced hardware is the play here.
For €350 you have a vacuum with home station, with mop, with carpet detection and lift function and with proper software.
I'd rather buy that and change some components to have local software.
Similar to what this hobbyist sells for home assistant compatible home ventilation: https://github.com/arjenhiemstra/ithowifi
There's much to criticize about the dev, but there's really no way to make it significantly easier. Most robovac companies just don't want you to flash the firmware.
The dev opposes selling the connector PCBs, but people have always ignored that and sold them online. They're not hard to find, but having the PCB is really only the first step.
Sounds very reasonable - I don't have a robot vac at the moment because there are some very large thresholds in my flat and the floor itself in places is very wobbly, but if I had a guide to safely mod one to work for my residence I'd definitely have a go.
> ...it's massively expensive to buy the components piecemeal.
I think the point of this is not to make "a product". It's just a fun project that you can build yourself or participate in some way with it's creation and/or funding.
I think about parallels to custom auto hobbyists. There are a lot of shops that can do incredible builds matching up different parts, engines, suspensions and even large scale chassis mods, but there are comparatively few outfits setup to build the car from scratch. A car is a different scale endeavor than a vacuum but some of the tradeoffs are similar. Maybe the engine, transmission etc mechanical interfaces are more easily modified and adapted than digital electrical software interfaces
This brings up the question: Where does one go to get a white label vac? Say I want to start a robot vacuum company and write my own custom software for it. How do I get a few white label samples from a manufacturer to do that?
Dreame's rebranded Mova starts at something like 350 EUR. Yes it's kinda capable and yet still kinda shit - gets tangled, stuck and needs quite some TLC. It doesn't look like it's going to be very reliable either.
I can't imagine how poorly 70 one can be.
Reminds me of people buying battery stick vacs without checking and then getting disappointed it's not same as dyson (while samsung as actually leads according to project farm tests).
Maybe. At the very budget end, I was really thinking you take the £150 worth of sensors and motors (when bought new) to make something better.
We just got a second hand Ultenic T10. The 2021 model is commonly available for £40 used. It tangles a couple of times a week and its battery will probably need replacing eventually, but it maps well and empties itself. At that sort of featureset, just a brain transplant to get it offline would be a welcome upgrade.
> oomwoo is organized so the community can build it in parallel. The robot and its software are split into self-contained modules. You pick whatever module interests you, work on it whenever you want, and submit your work as a pull request. Multiple people can tackle the same module — the best solution surfaces over time.
I think one major advantage of open source over commercial alternatives, is the possibility of endless improvements. Similarly, 3d printing as a manufacturing method allows for a short iterative cycle, high degree of design freedom, customisation as a product feature, local production, and an high degree of repairability.
It’s going to be interesting to see how well git(hub) and discord serves as collaborative tools in this case. Hardware files are often binary, hardware components have complex interfaces between them, and hence depends more on human communication and collaboration.
I really hope this project succeeds. I’d love a cloud free robot vacuum that I can trust.
Even though it’s vibe coded, I like the idea of an open source repairable robot vacuum. The current generation of them are notoriously not built to last / be repairable.
I don't know why you would say that, my Xiaomi s6's wheel motor died, I was bummed about it. I ordered a replacement motor, and to my surprise, I only had to open one or two screws and the motor module popped right out. The module had a nice slitting connector. I put the new motor back in and I was done. The thing must be at least 8 years old by now and it's still chugging along. I now passed it on to my parents and it's cleaning their house.
Agreed. I have a 6.5 year old Roborock S5 Max, and it still works fine. I've replaced a few parts (can still get on AliExpress), but other than that no issues. It's cleaned 74km2.
Also very fond of the Roborock S5, in fact I recently got a second one for the other floor - totally took it apart, cleaned it, put it back together and stuck Valetudo onto it.
The first one is from 2017 and still going strong - issues so far: battery replaced (only recently), laser motor replaced, fuse replaced. Aside from the fuse it was very easy and doable for basically anyone.
I’ve had two S5s die on me recently. They kept shutting down in the middle of a clean, and from what I read, it needed a battery replacement.
Ordered one off Ali Express, and after another couple months, it also started dying. So replaced it with a newer Roborock.
Didn’t bother when the second S5 started doing the same, just got a new Roborock.
Both new ones have been going well so far, and while it does seem to be good for replacing parts (I had another lidar part fail, and the replacement was easy), I was disappointed that replacing the battery didn’t fix the shut down issues.
Same here: 10 year old Roomba from the 6xx series, still going strong. I bought official replacement parts for a wheel, some brushes and a new battery: Replacing them was very easy - just a few screws, no glued-together parts.
Assuming you live in a crazy big house (1000 m2), it cleaned your house 74000 times. Given S5 was released 8 years ago, even approximating it to 10, that's 20 times a day. I can imagine it taking about an hour to do one clean run. Jesus, did it ever stop cleaning
no way it covers 1000m2 in an hour. It is ~35cm in diameter, so it has to travel around 3km. That would mean it'd require around 1m/s, which is 3times its rated speed... Then of course, it'd have time to charge. Even with 1C it's still 1 hour between cleaning sessions. We assume regular battery changes every couple of weeks too
yup, Xiaomi products are generally easy to repair. I've replaced a suction motor on a Roborock, built one working electric scooter out of two broken (in direct ways) ones. The firmware on the scooter is easily replaceable, the one in the vacuum makes it easy to install valetudo. If only more manufacturers were this way.
Agreed. I can code so I don't care whether it's vibecoded or whatever to bootstrap. Them working on designing hardware is what matters to me. I'll definitely keep an eye out for the kit, I don't have a lot of patience for hunting parts but would love to play with this.
The issue I have is the documentation and “status” is slop. Looking at the repo, how much of it is even real?
There’s supposed to be a build-along on YouTube but nothing there yet. The BoM is a bunch of aliexpress modules which is ok, but what about the chassis? Is that image generated?
The RFC calls to generate accurate models for the components, but the render looks like a full assembly?
When they get to the point of shipping a kit, why would I care? It's open source, just fix things. It's not rocket science, I'm just no good at working out the baseline machine that has parts to do the things. So I can't help at this point.
I don't expect a finished product. The value to me is the customizability and figuring out how to make it do what I want it to do. I'm sure that's not for everyone but like I have fingers. I can type. I can fix things. Slop is perfectly fine as a first draft because I'm envisioning a community of builders not a bunch of entitled twats who should just buy a Roomba.
A few thoughts on the vibe coding… This is probably just one person and this project wouldn’t have seen the light of day if they weren’t able to vibe code it. A few years ago this would have to be a kickstarter that raised at least several hundred thousand- probably millions to have a shot at successfully getting off the ground. You’re talking software and hardware engineering, experts in multiple disciplines, a whole team of people pouring in many hours to develop a product, etc.
Vibe coding doesn’t always have to result in low quality. An experienced engineer with good systems design skills piloting an agent can be incredibly productive. Although I’m pretty rusty at writing code, I’m still good at systems design and I’m having success with coding agents.
Recently, I’ve built a system for myself because what I wanted didn’t exist. There’s no way I ever would have done it without AI. I wouldn’t be able to pull it off myself even with years of time and a budget to hire developers for my personal project is nonexistent. It’s the kind of thing I never would have thought to start prior to good coding agents.
My productivity has been insane. I feel like there’s 10 of me. The quality of output is shockingly good. I’m looking at this and it’s one of the most put together systems I’ve worked on at any point in my career. It’s beyond what I saw from much more senior developers than I and it’s beyond what I was ever capable of myself.
I get why people don’t like vibe coding. It does produce a lot of slop in the hands of someone unskilled in the use of their tools. It costs people their jobs. There are a hundred reasons not to like it. The flip side is we get cool projects like this one because a single person can build the thing they always wanted and never could until now.
I've always wondered why the Robo mops have lagged so far behind the vacs, to the point that wirecutter (at least as of last year) wouldn't even recommend one. You'd think they'd be on par by now even with the added complexity of water. Pool robots deal with that quite well.
Personally, I find open hardware to be the selling point for devices that are supposedly running open source. If I can't change the parts/components, there's really no point.
As someone who works across software, hardware, and deep learning, this project makes me happy. My $1000+ vacuum will happily get stuck on the cloth mat, eat all our lego, and it has seen a hundred times, no memory, no meaningful object reasoning, just a marketing claim of intelligence propping up the price. This is something I will definitely play with.
The name is exciting to me. I've been a multiple time robot vacuum owner and it does have an appeal to be able to see a fresh build dissected like this. Why not contribute to this project instead of having a go all on my own, except of course with my AI helpers. I could pick the vacuum control board for the motors and sensors. I have some thoughts on brushes too.
It's a good point, vibe coding does lend itself to fast splitting among developers with the intent of recombining quickly too into a larger project.
I wonder if there's a project like there's OpenWRT for routers, just for vacuum robots? Where you just can buy some device form a normal online shop, then flash your custom firmware, and off you go with a privacy respecting fellow that doesn't film you and doesn't send everything to some US or Chinese cloud. :)
(I still use a traditional vacuum here, because all the privacy and snooping aspects of those robot thingies.)
The brushes usually don't go all the way to the sides of the vacuum, that's why many of them have an additional circular brush on the side to get into the corners and scoop the dirt to the big brushes in the middle.
No need for edges, also makes navigating through narrow things easier.
If you imagine a square robot traversing a wall and approaching a 90 degree inside corner, it can’t make the turn and would also be unable to make a perpendicular move to get more space.
That said Eufy, Makita and others make square-ish robots.
Mostly circular, some have slightly asymmetric shapes like Eufy e28. The mid range and higher typically have an extending arm that handles edges or corners
Interesting project. Aren't most modern robot vacuum's using image processing to determine whether to stop or not now though? How is Lidar going to help you avoid the cat's diarrhoea on the rug?
I am bone tired of slop. This looks like a useful thing to build (the cameras in existing closed source robo vacuums creep me out), but when people don't even write their announcement blog post by hand it gives me zero confidence in the project getting anywhere meaningful.
Perhaps not the place to share this, but it's depressing. I hope this proves me wrong.
It's a bit of an indicator about the effort they put into it. If they don't even write their blogpost themselves the question of "how much effort and thought did they put into the rest of their code / product".
Now, obviously they might just be bad at writing blogposts but surprisingly often it seems to be a decent red flag.
Because the thing is that the less effort you put into that the more anyone can just...reproduce the idea with their own LLM.
Even if s.o buil a cool thing and wants to share it with the world, if all they did was prompt Claude for a weekend what is stopping me from just doing it myself? Then I can even get it however I want.
I disagree. Writing became seriously unfun after the 8th year of school. Now after grad school it seems very wasteful to spend any calories focusing on the packaging of the ideas and things. Especially at a time when we have achieved the technology to not need a team of writers, editors, and typesetters to create presentations of a worthwhile quality for the actual content in a reasonable timeframe.
If there is a choice, I would much rather someone use an AI to do the project, and then used their human words to explain what the project is, rather than vice versa.
The critical difference is that a project artifact (software or mechanical design) is good as long as it works. It might not be maintanable, or editable or extendable, but it might narrowly just work. But explanations don't work like that. The content, the actual words matter just as much as the overall message.
An explanation can be thought of as software executed in your brain as you read it. I don't want to execute badly written software in my brain.
Of all the things LLMs do, one of the most fun is that they help you get over that hump of activation energy for an idea. We all have limited spare time, and going from "hey this might be cool" to a working prototype in minutes instead of days is intoxicating.
Much like my own heaving ~/prototypes folder, there is an avalanche of small projects other people are building in their own spare time (with LLMs), and there is a subsequent avalanche of "check out my cool project" posts. This is cool! However, unfortunately, almost universally, there is very little follow through. If you come back to those projects after a month, most are abandoned.
The creators of the ones that tend to last, at least in my brief experience so far, _do_ write useful blog posts by hand, or put a bit of human effort into sharing what they've built. I guess when I encounter someone sharing their work by way of blog post, it feels to me like they don't really care about actually sharing that work.
Also -- and this is much more a me thing -- I'm just fucking tired of reading Claude's writing. I have to work with Claude most days, and seeing it take over the whole internet is suffocating. Inflicting more of it on others just sucks.
my interest in this would be to customize the cleaning plan/logic... especially regarding traversing difficult carpets edges (which my roborock struggles with)
maybe something like:
- this specific part of the carpet is the best place to enter onto it
- once successfully on a carpet, stay on it until done cleaning it
It could still be the UI for this, if they wanted to, and they just implement the guts behind it with that in mind. It's not everything, but the UI is still something you can ~skip.
Heh that was my first thought upon seeing this headline: how will the Valetudo guy contort himself to get upset about this to the tune of an unhinged screed this time?
You can purchase a lidar vac for £70-80 now. Even if you only replaced the brains, that's a quarter of the price of a Oomwoo. The only upgrade I'd want is self-emptying. You'd probably have to relocate the charging contacts but it seems highly achievable.
Or you could break up an existing vac for the parts. You'd get the lidar, bumper, ToF, cliff sensors, motors and wheels, perhaps even some seals for your printed parts. Again, much cheaper, especially if you shop on the used market (I can get a whole working vac for the price of new wheels). All these robots use common parts so the risk of getting it wrong is very small.
My point is perhaps they could coalesce around a common white label option unit or set of parts currently sold as a vacuum.
The kit is a control board for the pump and boiler, and some add-on sensors for temperature and pressure. The "high end" features that it enables are almost entirely software driven, the main one being temperature control via PID. I've seen even simpler mods for other machines that bypass the "brew" button so you can do things like connect a bluetooth scale to enable brew-by-weight on a machine that doesn't support it, or add a shot timer.
The commercial version of this would be the Decent, but it costs 3x as much. I would love something like this for my robot vacuums. Valetudo is minimally invasive, but there's no reason you couldn't control the vacuum + wheels, but navigation is hard and those sensors are much more complex (can you even access the camera and undistort the image?)
https://gaggiuino.github.io/#/
† they pulled the rug on open firmware
They're fine to DIY if you're sensible about it. If you can build it, you can build a simple hydrostatic pressure tester. Pump, valves, tubes and a gauge. No computer parts. Pump the tank up to 1.5x rated pressure. This is industry standard.
But of course the age old saying applies: know what you're doing.
Edit: Ah, just found an "out of the loop" explanation here... https://old.reddit.com/r/gaggiaclassic/comments/1hbnd8r/out_...
I'd rather buy that and change some components to have local software. Similar to what this hobbyist sells for home assistant compatible home ventilation: https://github.com/arjenhiemstra/ithowifi
The dev opposes selling the connector PCBs, but people have always ignored that and sold them online. They're not hard to find, but having the PCB is really only the first step.
https://valetudo.cloud/pages/general/supported-robots/#eurek...
I did this, it's as easy as it is described. Fully functioning robot vacuum free from the cloud.
I think the point of this is not to make "a product". It's just a fun project that you can build yourself or participate in some way with it's creation and/or funding.
It's not practical. That's OK.
Dreame's rebranded Mova starts at something like 350 EUR. Yes it's kinda capable and yet still kinda shit - gets tangled, stuck and needs quite some TLC. It doesn't look like it's going to be very reliable either.
I can't imagine how poorly 70 one can be.
Reminds me of people buying battery stick vacs without checking and then getting disappointed it's not same as dyson (while samsung as actually leads according to project farm tests).
We just got a second hand Ultenic T10. The 2021 model is commonly available for £40 used. It tangles a couple of times a week and its battery will probably need replacing eventually, but it maps well and empties itself. At that sort of featureset, just a brain transplant to get it offline would be a welcome upgrade.
I think one major advantage of open source over commercial alternatives, is the possibility of endless improvements. Similarly, 3d printing as a manufacturing method allows for a short iterative cycle, high degree of design freedom, customisation as a product feature, local production, and an high degree of repairability.
It’s going to be interesting to see how well git(hub) and discord serves as collaborative tools in this case. Hardware files are often binary, hardware components have complex interfaces between them, and hence depends more on human communication and collaboration.
I really hope this project succeeds. I’d love a cloud free robot vacuum that I can trust.
Ordered one off Ali Express, and after another couple months, it also started dying. So replaced it with a newer Roborock.
Didn’t bother when the second S5 started doing the same, just got a new Roborock.
Both new ones have been going well so far, and while it does seem to be good for replacing parts (I had another lidar part fail, and the replacement was easy), I was disappointed that replacing the battery didn’t fix the shut down issues.
There’s supposed to be a build-along on YouTube but nothing there yet. The BoM is a bunch of aliexpress modules which is ok, but what about the chassis? Is that image generated?
The RFC calls to generate accurate models for the components, but the render looks like a full assembly?
Mmm I love the smell of slop in the morning
I don't expect a finished product. The value to me is the customizability and figuring out how to make it do what I want it to do. I'm sure that's not for everyone but like I have fingers. I can type. I can fix things. Slop is perfectly fine as a first draft because I'm envisioning a community of builders not a bunch of entitled twats who should just buy a Roomba.
Vibe coding doesn’t always have to result in low quality. An experienced engineer with good systems design skills piloting an agent can be incredibly productive. Although I’m pretty rusty at writing code, I’m still good at systems design and I’m having success with coding agents.
Recently, I’ve built a system for myself because what I wanted didn’t exist. There’s no way I ever would have done it without AI. I wouldn’t be able to pull it off myself even with years of time and a budget to hire developers for my personal project is nonexistent. It’s the kind of thing I never would have thought to start prior to good coding agents.
My productivity has been insane. I feel like there’s 10 of me. The quality of output is shockingly good. I’m looking at this and it’s one of the most put together systems I’ve worked on at any point in my career. It’s beyond what I saw from much more senior developers than I and it’s beyond what I was ever capable of myself.
I get why people don’t like vibe coding. It does produce a lot of slop in the hands of someone unskilled in the use of their tools. It costs people their jobs. There are a hundred reasons not to like it. The flip side is we get cool projects like this one because a single person can build the thing they always wanted and never could until now.
As best as I can tell, this project doesn't exist yet, just a bunch of boilerplate.
and
>It does produce a lot of slop in the hands of someone unskilled in the use of their tools.
Form a somewhat contradictory pair.
It's a good point, vibe coding does lend itself to fast splitting among developers with the intent of recombining quickly too into a larger project.
(I still use a traditional vacuum here, because all the privacy and snooping aspects of those robot thingies.)
No need for edges, also makes navigating through narrow things easier.
If you imagine a square robot traversing a wall and approaching a 90 degree inside corner, it can’t make the turn and would also be unable to make a perpendicular move to get more space.
That said Eufy, Makita and others make square-ish robots.
Perhaps not the place to share this, but it's depressing. I hope this proves me wrong.
Now, obviously they might just be bad at writing blogposts but surprisingly often it seems to be a decent red flag.
Because the thing is that the less effort you put into that the more anyone can just...reproduce the idea with their own LLM.
Even if s.o buil a cool thing and wants to share it with the world, if all they did was prompt Claude for a weekend what is stopping me from just doing it myself? Then I can even get it however I want.
The critical difference is that a project artifact (software or mechanical design) is good as long as it works. It might not be maintanable, or editable or extendable, but it might narrowly just work. But explanations don't work like that. The content, the actual words matter just as much as the overall message.
An explanation can be thought of as software executed in your brain as you read it. I don't want to execute badly written software in my brain.
Much like my own heaving ~/prototypes folder, there is an avalanche of small projects other people are building in their own spare time (with LLMs), and there is a subsequent avalanche of "check out my cool project" posts. This is cool! However, unfortunately, almost universally, there is very little follow through. If you come back to those projects after a month, most are abandoned.
The creators of the ones that tend to last, at least in my brief experience so far, _do_ write useful blog posts by hand, or put a bit of human effort into sharing what they've built. I guess when I encounter someone sharing their work by way of blog post, it feels to me like they don't really care about actually sharing that work.
Also -- and this is much more a me thing -- I'm just fucking tired of reading Claude's writing. I have to work with Claude most days, and seeing it take over the whole internet is suffocating. Inflicting more of it on others just sucks.
maybe something like:
It is not a robot firmware and has no role here.
Let alone why would someone want to attract the toxic culture that is the Valetudo creator and community?
This project seems like AI slop, but at this point that’s better than toxic dictators.
The AI slop on the site is not appealing, but it could also mean that the project will be parallelized successfully.