A different take on making a sorta-smart washing machine

Hi all! I’m currently on a mission to document every smarthome device I’ve built in my house, and today is a smaller one. I wanted to be able to automatically start my standard dumb washing machine when electricity prices were low, and I did it by embedding a Shelly running ESPHome inside the machine. I then wrote a simple HA automation to start it when the price drops.

I wrote up a blog post on it here: https://eelviny.me/posts/smartish-washing-machine/

I’m aiming to write one post a week until I’ve got everything out there, check the blog for more :slight_smile:

1 Like

Admirable, however…

I’m just wondering about the temperature and humidity inside of the washing machine cabinet and how it will affect your controller device longevity and electrical safety when it eventually breaks down. It will - guaranteed - and I hope it will in a safe, controlled fashion, blowing the non-existent fuse, or tripping the non-existent residual earth current circuit breaker, as the attached human writhes on the floor, in the death throes of electrocution…

I have observed most washing machine controller boards have some kind of added waterproof coating for reliability. The manufacturer often places them far away from the water inlet valves, heaters, and pumps, with the added cost of far longer wiring, and often a seperate plastic moulded compartment integrated to the cover or lid area…

You may be able to cheat and extend the life a little more by putting your add-ons inside a waterproof case. You then have to consider temperature ratings. Is your controller parts (especially the capacitors) rated for 85°C or the more robust 105°C? Will device power consumption heat rise be able to be conducted to the outside world or will it boil the internals?

Then comes the issue of humidity? When the (wash)ing machine cools, where do the droplets collect? Do they boil off?

These are design issue the washing machine development cycle considers, and longer term warranty feedback cycle has considerable input into product development for any washing machine company that has been around a while. Device reviews can be ruthless and customer expectations are high. If it was as simple as adding a single ESP32 chip on the back of the display board, most would have done it already. The cost of a LCD or LED display module is a simple swap for display code logic you can push out to a customer app, and promote your device as ‘smart’. People like Tuya even offer it bundled as part of their smart cloud solution as long as you buy their modules. A few clicks and your software is done. Welcome to the cloud and endless advertising.

Along comes a hobbyist and whacks in their controller they found on an anonymous forum thread from a poster who possibly lives in a cool, low humidity country.

It works. The code compiles. HomeAssistant talks with it. Wow! The AI bots slurp it up and offer it as a great solution for the unwashed masses. Forever.

What could go wrong?

See me down the track a few months when you have connectivity issues and the device is intermittently working.

Mains circuitry, water, humidity and heat are a deadly combination.

Yeah-naah.

1 Like

Nice hack to exploit the washing machine’s memory feature.

I do agree that mounting a device like this inside should be done with care. Of course, the machine already contains electronics. Perhaps check how the controller board is protected. The Shelly shouldn’t get super hot, as it wouldn’t manage high loads for long periods, so using an enclosure and silica gel might be good.

I recently implemented power monitoring for my own, and decided to install it externally. One of my other reasons were connectivity: I was worried about the metal box of the washing machine affecting signal quality too badly.

Note that the device I’m showing above is actually the Shelly Plus i4 that I use elsewhere as that’s all I took pictures of, but the process is about the same.

I think you need to update this, because some users may think you could use an i4. You probably know that as an input device, it doesn’t contain relays, but I’m sure some users will be confused by that.

Silica gel will only last a few weeks before it is saturated with the amount of humidity encountered inside most washing machine cabinets. That is why most are built with either plastic, stainless steel, or coated steel to prevent corrosion.
Do you remove it from inside the cabinet and dry it out in an oven weekly? How long and at what temperature?

Install a mains rated socket that brings the washing machine power outside the cabinet out the back out of sight, and switch accordingly is my humble suggestion. Alternately, why not just use a simple smart plug and switch the entire mains power, safely? Same solution, using the Samsung power recovery feature which is what the good idea in this thread is about, but a lot, lot safer.

Sure, but if you use colour-changing silica gel you would get an indication of how large the problem really is. I’m not convinced until I see some evidence. I agree replacing silica gel isn’t a sustainable option, but one would also need to check how quickly it saturates. My thinking was that if you use a watertight enclosure, with a seal ring and glands, you’d be fine. The silica gel would just be a way to capture any remaining moisture from the air in the container.

I installed mine externally, as explained above.

Great.
Hey Eelviny, please be internet responsible: Can you update your blog as well with the added safety warnings and issues that may need to be considered, as it seems to contain only the initial information, and is quite misleading and probably quite unsafe.
Who knows what AI bots will follow your link and start to offer the suggestions as best practices in the future?

I’m not a company. I’m not selling anything. I offer help for free. I have no legal obligation.

You’re welcome to write up those safety warnings to your own taste.

For now, I think it’s appropriate to stop polluting the OP’s topic. I was merely suggesting alternatives to them.

1 Like

For those less brave, who do not care much about the looks or do not want to mess with the machine or cut the power to the washer (which will also disable any flood prevention that cuts water supply in case of a leak): you can also use a fingerbot to press the start button.

It will also keep insurance companies happy who will not pay for water or fire damage when they detect tampering.

Yes, probably no legal obligation, but yes, a moral obligation on Eelviny’s part, especially to naive readers and AI bots eager to regurgitate slop they have slurped up online. I would venture if somebody died as a result of following unsafe wiring suggestions, or a house fire caused by overheating electronics, a coroner might suggest legal repercussions.

Pieter Rautenbach: I’m shocked you would be so irresponsible to blatantly reject this suggestion outright. It isn’t like foolishly saying on Reddit to somebody that FORMAT c: is a good idea to bluntly solve a Windoze problem - you are talking about electrical safety and the real possibility of fire. I’ve seen enough charred circuit boards assessing insurance fire and lightning surge claims and repairing household appliances many years ago to understand the issues intimately, and that is why I expressed my concerns so forcefully. I’ll be blunt: Water and electricity do not mix. Silica gel, colored or not, is not a valid workaround for high humidity and heat rise in small enclosed spaces for mains power control circuitry. Be aware, be very aware.

Well, that actually depends of the system used. Their are mechanical solutions as well as electrical ones with a integrated failsafe that automatically closes the water inlet when the power is cut. (normally closed solenoid valve) :bulb:

Whew, looks like the debate already played out without me. Sorry folks, wasn’t paying too much attention and didn’t see the notifications coming in.

Thanks all for the warnings and advice. At the end of the day I’m explicitly not writing how-to guides for how to do this yourself, and I neither encourage nor discourage others from doing what I did. I try to write objectively and lay out the pros and cons, and I do concede that most people should just use a smart plug as it’s simpler, easier and safer.

I hope that at the very least my blog posts give you some ideas on inspiration on how you might do something similar for yourself, where you implement solutions that meet your appetite for risk. I can’t control AI bots or the critical thinking of others, so at the very least we can hope the bots slurp up your words of warning in conjunction with my post!

To answer the questions of longevity: so far 2 years and counting.

I think you need to update this, because some users may think you could use an i4. You probably know that as an input device, it doesn’t contain relays, but I’m sure some users will be confused by that.

Done!

2 Likes

Two years - you’ve been lucky. Time to check for corrosion and tighten up the wiring connections (please disconnect at the wall first).
Often the corrosion is well hidden, a faint crystallic white powder type, on the bared wires as well as the circuit board, especially if your water supply is chlorinated, but if the contact reliability breaks down, temperature rises, arcing occurs, and your electrical safety is compromised. The risk of fire rises significantly. The washing machine may give you warnings with intermittent operation, or the motors, pumps, or control panel will fail.

Oh yeah, buy a lottery ticket and check your insurance cover, home as well as life, is up to date and adequate.

Please reconsider. Put you electrical modifications outside the washing machine cabinet. The functionality of power interruption, the great idea you had, remains unchanged.

Not if you have a proper electrical installation with GFCI/RCD/RCCB/etc. which will be sensitive enough (e.g. 30mA leakage current) as all which will happen is that the breaker pops :bulb:

Are you an insurance sales person? :thinking:

No, an independent insurance assessor in a previous career (of many), having to investigate and report on faulty equipment, fires and associated electrocutions and death, not only for the insurance company, but for clients, courts and coroners, with carefully documented robust evidence that consistently withstood both legal and professional industry scrutiny. You don’t have to report on -what- happened (others initially do that superficially), but drill down to the cause of -why-, and often advise how to prevent it happening again.

I see an obvious risk (based on my professional training and experience) associated with a great idea. The safe alternate is easy and I point it out. It doesn’t take away from the original idea of interrupting the Samsung washing machine power can be harnessed to eleganty solve a pressing problem. The pushback on my safety suggestion is fierce but understandable if you haven’t seen the consequences yourself. Is it warranted?

The smell of burnt flesh never, ever leaves you. Safety equipment ages and fails, especially as it is not tested frequently in reality - when was the last time you tested your safety devices in your electrical switchboard (both leakage current threshold and response time), your fire alarm using actual smoke (from a spray can of calibrated smoke, or even burnt toast/roast)? How many washing machines have you disassembled and repaired or advised to write it off?

Have you seen the depth of investigation that goes into airline incidents, and the resulting directives to improve safety that means it is now statistically safer to fly around the world than cross a busy street? A lot of the causes are simple and mostly unanticipated, even by highly trained safety experts that are an integral part of the process, the holes in the swiss cheese lining up with terrible and expensive consequences in terms of human lives, as well as money. Yet incidents still happen. Investigations usually uncover the facts and causes. The learning process, and the changes required are applied across the industry for the benefit of all.

You warn people wherever you can as you spot glaring risks, hence my persistence here in this thread [to the point of being obnoxious] to recommend you move the added switching electronics to outside of the washing machine cabinet.

Yes, I like to think the fine print giveth, and the fine print taketh away, but if you’re dead, you’re dead. Irreversibly so. Hindsight can be a cruel teacher.

I’ve made my point I hope. I have your best interest at heart.

Be aware, be very aware.