Have a fine mesh on a 45 degree angle that will allow the whole beans to slide down the mesh, but that water will pass straight through. The beans slide into a permanently covered bean receptacle that is impossible to add water to without disassembly (which should be an option for cleaning).
Beneath the mesh you could have a catchment to collect small bean particles and any water that the user tries to add, and the design and placement should be quite obvious that water poured onto the mesh isn't going anywhere. Conversely, let it collect there and go nowhere. If that catchment overflows, it goes onto the counter.
This was proposed, but especially if the bean capacity is large getting the beans wet is a big no-no. It also does nothing to dissuade users from adding more water for whatever reason (until it overflows). Thus you get potentially moldy beans, and a tray full of gross old water.
Only a serious idiot would ever try to pour water on that. The beans slide down the wire mesh into the main bean compartment which is otherwise inaccessible.
Some machines have a wire filter like that to stop objects getting into the water tank; users might misinterpret it. I still really prefer the ones where you have to take out a basket, which is obviously not capable of holding water, and fill it with beans. Sealing it and getting the beans is harder.
For a user to misinterpret and pour water on what biot described would require such a fundamental misunderstanding of common sense and what I can only hesitatingly call "physics", that I don't think it is possible to protect the machine from such a user - except, perhaps, to hope that they will be foiled by the doorknob whose operation they must decipher in order to enter the break room.
The mesh wouldn't need to be that fine if it just had to exclude whole coffee beans.
I think I'd use a bottomless bean bin that pulls out with a handle. While it's pulled out, it's over a grid that lets water pass through (and ideally makes it clear that any water added is just going to be dumped right on the counter). When it's pushed back in, the beans fall into the actual bean reservoir.
The key advantage of the mesh is not to perfectly prevent water from flowing along it, but to be something that even the most caffeine-deprived user would not accidentally try to pour water on. Even if it does nothing mechanically to prevent water from entering the machine, it will psychologically prevent misuse.
I worked at a place where someone (who we never caught) put some chocolate coffee-bean-look-alike candy from Trader Joes in the grinder. All the new peeps (myself included) got 'blamed' for being ignorant, and lots of signs went up. I hate signs because the can never anticipate all scenarios.
My next work place had someone try to cook bacon in the toaster oven. As it turns out, this is not a viable way of frying bacon and it caught the toaster and part of a cabinet on fire. Later that month the building management started doing regular fire drills. We'd take the opportunity to tell the story about Mike and his bacon as we stood on the street.
I'd say people were idiots, but that's a negative view practiced ineffectively by the first place. I'd rather see it more like people incorrectly assume what they are doing is right, and never question if its not. Silly humans.
Signs are also stupid because they're built on two faulty assumptions: one, that the morons read them, and two, that the mistakes of the future will be carbon copies of mistakes of the past. There's no bacon frying sign over the toaster oven at my work, and I've seen people cook bacon in the kitchen. But there's a sign over the sink listing things that shouldn't go down it, as if nobody has one of these things at home.
My previous workplace had someone attempt to cook a ribeye steak in the toaster oven.
It actually worked, but management was angry the third time he pulled it and set a flat no-meat-cooking policy on the toaster oven (enforced via clipart signs).
I cook bacon in the toaster oven all the time... it's the best way. As long as your toaster oven can maintain 425F / 215C constant temperature (rather than having the element on permanently) you end up with beautiful moist but cooked bacon without it being crispy and burnt.
We had a coffee machine where you had 2 containers, one for the beans and one for the water, never had a problem with knowing which was which. But there was also a place to put in pre-ground coffee, which we never used.
One day a consultant came along, and not being able to make coffee because the machine had run out of beans, they proceeded to pour beans into the space reserved for the pre-ground coffee. We found this out by the horrific gear crunching noise the machine had made as the coffee beans shredded the internals to bits. It was a disappointing and sad day for the engineers, many of whom were frequent coffee drinkers.
Needless to say after the machine was repaired we taped that pre-ground coffee access port closed.
Use a padlock. Put the key where only knowledgable users can find it. This will cut down on your casual visitor screwing things up.
Engineer a dyson-vacuum-cleaner-like apparatus which 'sucks' water and beans in different ways.
Do the first comment's ideas, but make it a two-stage event - first is to put the item on the grid/flap, second is to manually fold the gridflap into the machine, where it's dumped into a temporarily unsealed beanbox. Water won't stay in the grid long enough to hang around for the second stage.
Create a sniffer that requires the scent of a fresh bean to be presented before opening the port.
Drink hot chocolate instead
Sell special iron-laced beans and use a magnetic apparatus.
A label saying "This container is for coffee _ _ _ _ s" with a lettered padlock (http://wordlock.com/padlocks/) would perhaps be preferable to the hidden key solution.
All you need is to design your bean enclosure so it's never completely empty. The few beans trapped in what I'd call a "reminder reservoir" will be enough to remind users beans go here.
I'm going to bring up the Microsoft Starbucks machine again, as it has also solved this problem. They used the 2nd solution - a water line directly to the machine. Also, the beans-hopper holds 2 kinds of beans, usually house blend and decaf, which rarely go empty at the same time.
No worries about watering the hot chocolate powder hopper, either. says "Hot Chocolate" on the compartment lid, and inside it's covered in chocolate dust.
I'm not sure if you've seen one of these coffee machines. When you come to one, your first thought isn't "where do I put water", it's "what kind of coffee do I want?"
The article already has a picture and it is a bit different than that.
Even with machines where you have to put the water in you don't have to do it for every cup, most likely people either put water when they put in beans, because that's what they are used to, or perhaps when the machine isn't working properly and they are trying to get it going again.
Don't have any opening to fill in beans or by mistake water, Get beans in a sealed container, sealed with cardboard at the bottom, which the machine can penetrate. Another solution is to connect bean jar through a pipe to bigger source of beans similar to how water is sourced. The bean source is maintained by a dedicated person. If they cannot learn how at use it,restrict access thats the way to go!
Simple elegant and effective. If it takes one second to transfer water from sink to machine, you only get one second to process the data and possibly get the wrong conclusion. If it takes 60 seconds to walk across the room, that provides time to think ... why did they make it so hard to add water to this thing? Plenty of time for data processing to kick in and notice ur doin it wrong
This may sound dumb, but maybe the easiest solution is to put a transparent fake water container that never drains. So anyone who is about to pour water would see that the machine already has water.
Since the machine in question is hooked up directly to the tap, one never needs to fill it with water.
That's what I was thinking, but then the water would stagnate, go green, and be a potential source of mosquitos. So someone will want to change it.
The fix would obviously be to make a little bit of the water be drained away whenever a coffee was made, but then it would need refilling and you're back to a situation where you might as well just have 2 containers and disconnected from the water supply.
Since the false water container on top isn't going to be used anyway, what about just filling it with something hostile to microbes, like dilute hydrogen peroxide? If you wanted to get really fancy, you could even add a couple lines marked "max" and "min" and add some plumbing to make the fluid level vary between them.
Apart from the problem of having potentially toxic substances as part of a beverage generation device, adding extra moving parts to simulate a non-moving part is really adding unnecessary complexity, effectively trying to fight against the universe (and the idiots that it produces).
Instead of a liquid, it could be filled with solid plastic, similar to the fake glasses of liquid you see in joke shops.
And perhaps a subtle label on the surface, for someone who notices that the container does not really have sloshing-around liquid in it: "No water required"
I might be wrong, but it's just water, I don't think it can go bad unless it's got bad things in it. And keeping it sealed would keep away the mosquitoes. If that ever even becomes an issue.
The solutions that include an open air design for the bean hopper ignore the fact that oxygen is what causes beans to go stale. This is why coffee bean bags have a one-way valve built in: to let the CO2 out that is given off by the beans, and to prevent O2 from seeping back in.
And the solution to the problem is to install a barista.
Eliminate the bean compartment and dispense the beans from a centralized location just like the water. Either using an augur or a pneumatic tube delivery system.
We have a cafe style coffee grinder (next to our manual espresso machine) that has two hoppers: one for the beans, and one where the ground beans end up, from which you dispense into the porta filter.
I am guilty of filling the wrong hopper with beans (talking instead of paying attention). Twice. Not quite as bad as water, but still a PITA to pour everything out and separate the beans from the ground coffee.
The coffee snob in me wants to say that for a nasty automatic machine like that, putting water in the bean hopper and breaking it is probably a good thing! Seriously, used commercial machines are fairly cheap and will make a far superior coffee which is worth twenty seconds of work.
But if you must, I like the idea of a bean container that must be removed to be filled. Better than warning labels (that people will ignore) or putting a picture of a bean there or something.
We have a proper machine in the office and some people still drink ugh instant. At least with a machine like we are discussing people will drink a better standard of coffee than they otherwise would.
I actually know people that prefer the taste of instant. I think it's one of those things where if you're raised on garbage, you'll like the taste. I know for a long time I though Tim Hortons coffee was the tits.
Load the beans from underneath using a detachable tray with a mesh inlay which will is used to pull the beans in.
(If you must go all out, put a bob that opens a trap in the bottom of the tray so they can't get the water near the machine if they try to fill the tray with it.)
I was going to post something on their site, but apparently you have to have enough karma to give a response now.
Use a removable side-loading bean cartridge. The bottom of the cartridge would be mesh, the top open. There would be a spring-loaded door on the side of the cartridge that the machine would open once inserted (or it could be opened via a "lock" lever for a pure mechanical solution) so that beans could flow out. This way it could never be loaded with water, and the beans would stay fresh since once inserted the cartridge would be sealed on top and bottom by the rest of the machine.
“This question is protected to prevent "thanks!", "me too!", or spam answers by new users. To answer it, you must have earned at least 10 reputation on this site.”
It's basically like semi-protection on Wikipedia: things which have gotten a lot of attention can get lots of noise, so only more established users are allowed to add content.
This is full of very serious answers. I thought this was a rhetorical question to highlight the tendency of software developers to mismanage risk.
I feel like no matter what the answer is, the questioner could just throw another "But what if..." on top of it.
Set a post-it note saying "Water here".
But what if a Black Hat gains access to the coffee machine and moves the sign?
Etc. You can't win unless you know what the questioner wanted to hear. I guess that makes it a good interview question.
Adding to that, regarding a comment on that thread: "It'll probably still be gotten wrong by the same people who push on 'pull' doors". Culture is very important. In my native language, 'push' actually means pull! I will always get that wrong unless the design is good, meaning there's only one handle and it's on the side you actually need to pull. If there's an handle and the word 'push' just by its side than I'm going to pull it no matter what.
It's just that there is a lot of thought wasted solving a derivative problem that could be avoided by continuing to use a simpler solution to the first order problem. This is classical overengineering. At the same time, any good solution to the coffee problem will likely have to find a tuned balance between a number of traits and they will be different for each person.
There are certain abstract tradeoffs that aren't taken into account in the OP (if I remember correctly). In the case of the keurig machine, it solves the complexity and convenience by narrowing flexibility, increasing expense and maybe sacrificing taste (which are all completely valid tradeoffs). The french press or a ceramic dripper maintain the flexibility, expense, taste and minimize complexity at the cost of convenience.
92 comments
[ 4.8 ms ] story [ 189 ms ] threadYou deserve your problem if you use that terminology for whatever people are not reading.
Beneath the mesh you could have a catchment to collect small bean particles and any water that the user tries to add, and the design and placement should be quite obvious that water poured onto the mesh isn't going anywhere. Conversely, let it collect there and go nowhere. If that catchment overflows, it goes onto the counter.
http://upload.ecvv.com/upload/Product/200801/200631654352537...
Only a serious idiot would ever try to pour water on that. The beans slide down the wire mesh into the main bean compartment which is otherwise inaccessible.
I think I'd use a bottomless bean bin that pulls out with a handle. While it's pulled out, it's over a grid that lets water pass through (and ideally makes it clear that any water added is just going to be dumped right on the counter). When it's pushed back in, the beans fall into the actual bean reservoir.
My next work place had someone try to cook bacon in the toaster oven. As it turns out, this is not a viable way of frying bacon and it caught the toaster and part of a cabinet on fire. Later that month the building management started doing regular fire drills. We'd take the opportunity to tell the story about Mike and his bacon as we stood on the street.
I'd say people were idiots, but that's a negative view practiced ineffectively by the first place. I'd rather see it more like people incorrectly assume what they are doing is right, and never question if its not. Silly humans.
It actually worked, but management was angry the third time he pulled it and set a flat no-meat-cooking policy on the toaster oven (enforced via clipart signs).
We had a coffee machine where you had 2 containers, one for the beans and one for the water, never had a problem with knowing which was which. But there was also a place to put in pre-ground coffee, which we never used.
One day a consultant came along, and not being able to make coffee because the machine had run out of beans, they proceeded to pour beans into the space reserved for the pre-ground coffee. We found this out by the horrific gear crunching noise the machine had made as the coffee beans shredded the internals to bits. It was a disappointing and sad day for the engineers, many of whom were frequent coffee drinkers.
Needless to say after the machine was repaired we taped that pre-ground coffee access port closed.
Engineer a dyson-vacuum-cleaner-like apparatus which 'sucks' water and beans in different ways.
Do the first comment's ideas, but make it a two-stage event - first is to put the item on the grid/flap, second is to manually fold the gridflap into the machine, where it's dumped into a temporarily unsealed beanbox. Water won't stay in the grid long enough to hang around for the second stage.
Create a sniffer that requires the scent of a fresh bean to be presented before opening the port.
Drink hot chocolate instead
Sell special iron-laced beans and use a magnetic apparatus.
Even just gluing a coffee bean or two to the outside might work.
No worries about watering the hot chocolate powder hopper, either. says "Hot Chocolate" on the compartment lid, and inside it's covered in chocolate dust.
My theory: http://news.ycombinator.com/item?id=5175098
Here it is, to the right of a Keurig coffee pot: http://static3.businessinsider.com/image/4ef3bec5ecad04cd2c0...
Even with machines where you have to put the water in you don't have to do it for every cup, most likely people either put water when they put in beans, because that's what they are used to, or perhaps when the machine isn't working properly and they are trying to get it going again.
Since the machine in question is hooked up directly to the tap, one never needs to fill it with water.
Otherwise I'm e-mad
The fix would obviously be to make a little bit of the water be drained away whenever a coffee was made, but then it would need refilling and you're back to a situation where you might as well just have 2 containers and disconnected from the water supply.
And perhaps a subtle label on the surface, for someone who notices that the container does not really have sloshing-around liquid in it: "No water required"
And the solution to the problem is to install a barista.
if you see a coffee machine that has bean storage, it's almost guaranteed to suck.
I am guilty of filling the wrong hopper with beans (talking instead of paying attention). Twice. Not quite as bad as water, but still a PITA to pour everything out and separate the beans from the ground coffee.
But if you must, I like the idea of a bean container that must be removed to be filled. Better than warning labels (that people will ignore) or putting a picture of a bean there or something.
They probably just don't know how to operate the machine.
(If you must go all out, put a bob that opens a trap in the bottom of the tray so they can't get the water near the machine if they try to fill the tray with it.)
Use a removable side-loading bean cartridge. The bottom of the cartridge would be mesh, the top open. There would be a spring-loaded door on the side of the cartridge that the machine would open once inserted (or it could be opened via a "lock" lever for a pure mechanical solution) so that beans could flow out. This way it could never be loaded with water, and the beans would stay fresh since once inserted the cartridge would be sealed on top and bottom by the rest of the machine.
(Even the joke answers, such as adding a "coffee bean pipe", are already covered in the stackexchange thread.)
“This question is protected to prevent "thanks!", "me too!", or spam answers by new users. To answer it, you must have earned at least 10 reputation on this site.”
It's basically like semi-protection on Wikipedia: things which have gotten a lot of attention can get lots of noise, so only more established users are allowed to add content.
I feel like no matter what the answer is, the questioner could just throw another "But what if..." on top of it.
Set a post-it note saying "Water here". But what if a Black Hat gains access to the coffee machine and moves the sign? Etc. You can't win unless you know what the questioner wanted to hear. I guess that makes it a good interview question.
I usually fail at glass doors when PUSH and PULL labels are transparent so that I read "PUSH" and (reverse) "PULL" simultaneously.
X sounds like SH.
It's just that there is a lot of thought wasted solving a derivative problem that could be avoided by continuing to use a simpler solution to the first order problem. This is classical overengineering. At the same time, any good solution to the coffee problem will likely have to find a tuned balance between a number of traits and they will be different for each person.
There are certain abstract tradeoffs that aren't taken into account in the OP (if I remember correctly). In the case of the keurig machine, it solves the complexity and convenience by narrowing flexibility, increasing expense and maybe sacrificing taste (which are all completely valid tradeoffs). The french press or a ceramic dripper maintain the flexibility, expense, taste and minimize complexity at the cost of convenience.