Why don't auto sensing bathroom faucets/dispensers consistently work?

[jdraughn]

We've all seen the movie where the guy tries to get the faucet to come on by waving his hand, goes to a different faucet, and the first faucet comes on, then goes back to the first faucet and it goes back off. Funny stuff, but here at work I experience it too sometimes. Why does it seem like no one can make a reliable automatic faucet? The soap dispenser is also automatic and sometimes it takes like 10 seconds before it will finally trigger. Is it cold hands, room lighting, bad adjustments, or what?

[Paul Moir]

A couple things I've noticed:

Pulse solenoids are used to reduce power consumption in battery operated or battery backed faucets.  When activated, the current drop trips out the control circuit when the batteries are low or cheap.  The faucet forgets if it's on or off sometimes.

Poor technology abounds.  I replaced a very expensive Delta which I spent many hours messing with with a (unknown brand to me at the time) Sloan and haven't had a problem since.  It works perfectly every time, except when it's power supply is knocked out by accident and the backup batteries run down.  It's been about 5 years in an industrial setting.

[AndyC_772]

This sounds like a question which can be most effectively answered with a challenge. Design a detector for exactly this application, and see how it turns out.

I'm willing to bet it becomes apparent quite early on what the challenges will be with any given approach, and what the likely effect will be on reliability.

[Rerouter]

the cheapest way i can percieve to solve the issue would be a ultrasonic chirpier and receiver (i ruled out leds the more i thought about how do you tell a hand apart from the water flowing) fire off a very low amplitude shaped pulse (e.g. a two tone), and measure the returned peaks, if it changes drastically such as when an object is brought in the path turn on, then as we are playing with sound, start looking at peaks before the distance of the water flow to know how long to keep on, a hand generally intersect the flow, have a small timeout, e.g. 5 seconds, then turn back off,


At least that is my take on the problem seeing as low power micros with an ADC, and a peak detector circuit are easy to make,

wake up, chirp, if peak is within distance turn on, stay on chirping until peak no longer found, turn off, go to sleep

[tszaboo]

I think they are using IR reflection, and they just have too small sensitivity. I understand the issue, that the bathroom is not an ideal condition, and plummeting huge amounts of IR is not power friendly. But there must be a better way to do this. Like using a second occupation sensor to detect if someone actually stands in front of the faucet.
The ones I seen working awesomely are the ones at Charleroi airport. Too bad the design is moronic otherwise. They built the hand dryer and the faucet together.So that one has two sensors close to each other, one for the water, one for the air. The air goes to the sink, and if there is any water there that will be splashed into your face, and all over you, 5 minutes before your plane takes off.

[Delta]

These incredibly annoying things are a perfect example of the wrong technology for no problem.

I am a bit OCD about "germs", so what's wrong with a simple pedal on the floor to turn the tap on?

[Zero999]

I didn't know these things had batteries. I thought they were mains powered.

[mikecrowe]

The ones in the wash room where I work are powered from 2 AA's. For safety and convenience (pardon the pun) I'm thinking. Ultrasonics I think would get contaminated quickly due to the hostile environment.

[BillyD]

We have those at work and they seem to frequently get stuck in the On mode, or more often get stuck in an infinite On-Off cycle.
I suspect the latter is due to the warm water confusing the IR sensor.
The volume of wasted water must be staggering, not to mention the expense of purchasing such an expensive device to do a simple job.
A good example of a solution desperately looking for a problem.

[vk6zgo]

The Urinals in Singapore detect if someone stands in front of them,performs the customary activity,then steps away.
They then flush.

The weird thing is---the things know whether you peed or not!!

I tried the experiment of just standing in front of one then stepping away--no flush!

I only did this experiment twice---doing weird things in Public Toilets in foreign Countries may be hazardous!

[c4757p]

These incredibly annoying things are a perfect example of the wrong technology for no problem.

I am a bit OCD about "germs", so what's wrong with a simple pedal on the floor to turn the tap on?

Too simple and reliable. This sort of thing has to be poorly designed crap, otherwise you upset the natural order of convenience technology...

[Fat]

I hate those. I know enough to flush the toilet and can wash my hands myself. Even turn on the hand drier or pull down the next towel. Automated bathroom euipment is humiliating.

[Richard Crowley]

Here in my office, the taps appear to be capacitive detectors. If you touch any part of the chromed metal spout, it turns on just as if you put your hand under the water outlet.  The only real mode of failure is that they get progressive less sensitive to the point where you can't get them to turn on at all.

OTOH, the soap dispensers appear to be IR detectors. And their mode of failure isn't detection, but operational. The people who service the facilities use liquid soap of a much different viscosity than the pumps were designed for and many of the dispensers just make noises and dispense nothing.

I have thought of making a comedy short video of a zombie walking into one of these situations. Because of their low body temperature, the toilet/urinal won't flush, they can't get any water or soap at the sink, and the paper towel dispensers won't acknowledge their presence with a length of paper, either.  They are just non-persons.

[BillyD]

The Urinals in Singapore detect if someone stands in front of them,performs the customary activity,then steps away.
They then flush.

The weird thing is---the things know whether you peed or not!!

I tried the experiment of just standing in front of one then stepping away--no flush!

I only did this experiment twice---doing weird things in Public Toilets in foreign Countries may be hazardous!

We also have those at my work, and unlike the taps they work very well. In fact they probably compensate for the copious volume of water wasted by those bloody dopey automatic taps. They faithfully flush when you walk away after using them.
I've also been unable to fool them into flushing although admittedly I didn't try too hard either. They certainly don't respond to me putting my hand in front of the sensor. I think they seem to be able to distinguish this from the mass of a person standing about 18" away and probably also take timing into account.

[miguelvp]

The Urinals in Singapore detect if someone stands in front of them,performs the customary activity,then steps away.
They then flush.

The weird thing is---the things know whether you peed or not!!

I tried the experiment of just standing in front of one then stepping away--no flush!

I only did this experiment twice---doing weird things in Public Toilets in foreign Countries may be hazardous!

We also have those at my work, and unlike the taps they work very well. In fact they probably compensate for the copious volume of water wasted by those bloody dopey automatic taps. They faithfully flush when you walk away after using them.
I've also been unable to fool them into flushing although admittedly I didn't try too hard either. They certainly don't respond to me putting my hand in front of the sensor. I think they seem to be able to distinguish this from the mass of a person standing about 18" away and probably also take timing into account.

Urine is very conductive compared to just water. They just need to flush until they can't detect the conductivity.
Make a brine of water and salt and see if that flushes. Also try with just water and no salt, it should not flush.

[CatalinaWOW]

All of the ones I have examined used infrared sensing.  The windows are under the faucets and are subject to a lot of splashing, but are in an area where the cleaning staff (if there is one) will not usually notice dirt.  That would explain why the water tap controllers are much more problematic than the ones which control toilets. 

There is also wide variation in the sensitive zone, probably due to different brands and styles of plumbing.  I have found that many work reliably once you find the sweet spot for that configuration.  Maybe that would be a reason to use visible light instead of infrared.  A little modulation would easily sort the sensor light from the room lighting.

While it is easy to pick holes in the choice of infrared, I think most of the other choices will be problematic in that environment.  Capacitive sensors would depend greatly on installation by plumbers who are not the worlds best electronikers.  A stainless steel or cast iron sink would make the problem hopeless.  Ultrasound has its own set of problems.

[_Andrew_]

There is a simple way to have just about all of the taps coming on together and staying running which is waring a high visibility tabard. The reflective IR sensors react rely well to the light reflective strips on the tabard. 

[timb]

Here in my office, the taps appear to be capacitive detectors. If you touch any part of the chromed metal spout, it turns on just as if you put your hand under the water outlet.  The only real mode of failure is that they get progressive less sensitive to the point where you can't get them to turn on at all.

OTOH, the soap dispensers appear to be IR detectors. And their mode of failure isn't detection, but operational. The people who service the facilities use liquid soap of a much different viscosity than the pumps were designed for and many of the dispensers just make noises and dispense nothing.

I have thought of making a comedy short video of a zombie walking into one of these situations. Because of their low body temperature, the toilet/urinal won't flush, they can't get any water or soap at the sink, and the paper towel dispensers won't acknowledge their presence with a length of paper, either.  They are just non-persons.

Not Zombies, but I'll just leave this here:



"Manual Water Fountain (For Blacks)"

Damn, I really miss Better Off Ted. It was such a great show... Unfortunately, I think it was too smart for network TV. People just weren't ready for a new generation of Hurricane Proof Dogs.

[Paul Moir]

These incredibly annoying things are a perfect example of the wrong technology for no problem.

I am a bit OCD about "germs", so what's wrong with a simple pedal on the floor to turn the tap on?

Too simple and reliable. This sort of thing has to be poorly designed crap, otherwise you upset the natural order of convenience technology...

Requires special plumbing, extra hardware and (most importantly) makes it really difficult to clean the floor properly.  There are knee operated manual valves as well which solve some of the problems.  They're a little awkward to use though. 
The IR ones don't have to be poorly designed.  Some work very well.

[zapta]

Damn, I really miss Better Off Ted. It was such a great show... Unfortunately, I think it was too smart for network TV.

It's available on Netflix streaming. Great show, especially the Veridian Dynamic commercials, many available on yourtube.

[timb]

Damn, I really miss Better Off Ted. It was such a great show... Unfortunately, I think it was too smart for network TV.

It's available on Netflix streaming. Great show, especially the Veridian Dynamic commercials, many available on yourtube.

Yeah, actually quite a few of the Veridian Dynamic clips were uploaded by me.  It always surprised me that ABC never had them taken down, I guess they just didn't care enough...

Great show though. I totally recommend everyone check it out on Netflix Streaming; it may still be on Hulu as well. It's like Dilbert meets Arrested Development. (Complete with Portia de Rossi!)

[coppice]

The only place I remember seeing quirky behaviour from an automatic tap was the US. In Europe and Asia they just seem work, or occasionally be broken. Urinal flush sensing seems to be very reliable, too. Do US makers use different sensing techniques?

I'm not sure if automatic soap dispensers work badly, or are designed for humour value. They seem to wait just long enough for most people think they aren't working, and then squirt soap just as the hand is removed.

[station240]

I have the feeling the urinal ones have a float in the drain pipe as a sensor.
Fairly idiot proof technology that plumbers already know how to fix.

[Stonent]

Most of the ones I see are LDR or IR based. I suppose some of the issues could be caused by the flicker of the fluorescent lights in most bathrooms.