I love it. I used to work for a company targeting markets in the developing world. It's really easy to take for granted the supply chains that exist all around us. I always like to see the creative solutions people come up with when resources are constrained.
PS: As an example, note the sheet-metal construction. In an industrialized country we would laser-cut all these parts. If you wanted to make this in an area with less infrastructure you might use a template and carbide gas torch to cut out the large shapes, then a hand punch to make the screw holes. More labor intensive, but still doable.
In an "industrialized" country we encase the real drum in a glued tight unopenable plastic enclosure to prevent the drum seals from being service replaced when they fail so that what used to be a small low cost repair now results in a forced new machine purchase.
I have found that "low end" appliances avoid a lot of this nonsense since there's no money for extraneous parts and they've been using the exact same designs for decades.
Even on the low-end models, Whirlpool (and probably other brands as well) have managed to add at least one failure-prone and overly expensive feature:
The lid safety switch which prevents you from disabling the lid lock. It has a complex design with lots of anti-tamper circuitry. It's highly prone to failure and very expensive to replace compared to the price of the whole machine.
My thinking as well. If they're truly dedicated to making this available to communities everywhere, they should open the design and make it as simple as possible for everyone to build a washing machine at home. In many cases people would still buy them, which is fine.
> to make this in an area with less infrastructure you might use a template and carbide gas torch to cut
Metal work seems very expensive in some places. In a 2025 paper [1], a cooking pot looks like several aluminium rings welded together, about 50 cm tall and 70 cm in diameter, is 416 USD in Ghana, which is one of the destinations of this hand-crank washing machine.
I am interested in unit cost for mass production. It needs to be significantly cheaper than an old style top-loading washing machine to be affordable. The design of old style washing machine is mature and priced at around $100 for 8kg model. I suspect it can be stripped down further, remove water pump, remove program controlled inlet valve et al. to reduce the cost to below $50. Granted, washing machine like that needs electricity, but solar panel may be cheap enough.
One more thing, the water is not always easy to get in poor places. It is often much easier to carry laundry to a well, creak, or river than transport water to home. The path to the water sources may be a narrow trail often going up and down hills, so even with wheels on the machine, it is impractical to drag the machine to the water.
Extremely simple washing machines already exist, and I suspect on the order of 30-40 dollars. They are top-loading. No pumps. Turn one dial to let water in through the inlet. Turn another dial to let water out through the outlet valve. All manual, no pumps. Then flip switch to start spinning with electric motor, flip it back to stop spinning (no timers).
What you do is fill it with water. Add soap. Then put in first load of clothes and run it for 15 minutes. Then take out the clothes and put them in a tub. Repeat with second load of clothes in same soapy water. Once, all loads are done, then put in fresh water. Run all loads through it to get the soap out. You are done.
(Relatively) richer people might have another machine that acts as a spinner. Otherwise, you just hang up the wet clothes outside.
Of the stuff I've repaired recently it's been mechanical switches that have caused problems, where a microcontroller and bloody great MOSFET would have kept on forever.
There are much lover cost washing machines with electricity. This one for example: https://www.youtube.com/watch?v=2eo4CIHpp28 . It's just a motor, you just need a big bucket or a tub to fill it with water and detergent. Let it turn the clithes in water for half an hour and rinse. It's also handy during rinsing.
Growing up in a developing country in 90s we used to use this type of machine for bigger loads because our normal front loader machine was only 6kg capacity. This + a bathtub was the way to go for washing the blankets, bedsheets etc. It costs now sth like 20-30 usd.
A couple years ago somebody parked a stolen car in front of our building, stuffed a rag in the fuel filler, and lit it. I must have pulled up just a couple minutes after they had left. Grabbed the fire extinguisher I kept under the seat and put it out. I had that extinguisher around for about 5 years. You never know when you'll need it.
> From an embedded developer's perspective, Arduino is awful.
Specific AVR Arduino annoyances I remember:
* Strings loaded to RAM instead of program memory, so you use up all your RAM if you have a lot of text. Easily fixed with a macro
* serial.println blocks, so your whole program has to stop and wait for the string to be transmitted. Easily fixed with a buffer and ISR
* Floating-point used everywhere, because fuck you
* No printf(). It's in avr-libc, and it's easy plumbed in, but the first C/C++ function that everybody ever learned to use was somehow too complicated or something.
* A hacked-together preprocessor that concatenated everything, which meant you could only have your includes in one place, thus breaking perfectly good, portable code.
I think they ultimately did a disservice to novice programmers by giving them something that was almost a standard C++ environment, but just not quite.
Modern AVRs have program memory mapped into the RAM address space. The GCC linker scripts for the parts that support this put strings into .rodata within that memory region, obviating the need for special macros to retrieve them. However, you won't find this on most of the usual suspects in the Arduino AVR ecosystem.
> A hacked-together preprocessor that concatenated everything, which meant you could only have your includes in one place, thus breaking perfectly good, portable code.
While it does concatenate, does it break something regarding the include mechanism, assuming said files use include guards? You can include stuff wherever anyway (though doing it within a scope needs special considerations).
I learned how to program the AVR in assembly 25 years ago by reading the datasheet and various articles in Nuts and Volts. For its time the AVR had a very accessible development kit, the STK500, which cost about $100. A few years later avr-libc came along and if you were running linux and knew how to write C it was pretty easy to get started.
I'm sure somebody like me would happily take them off your hands. The AVR is still a solid platform for low-level applications. A lot of the Arduino libraries never really took full advantage of what you could do with that chip. Whatever happens with the Arduino IDE, those boards will still be useful tools for quite a while.
That's a funny take, since reading the history of Luddism I've always interpreted it to be anti-capitalist. The Luddites were peasants that had their way of making a living disrupted by a high capital expenditure technology that they could not compete against and thus took to destroying it.
reply