An Alarm Designer on How to Annoy People in the Most Effective Ways
atlasobscura.com
atlasobscura.com
https://www.youtube.com/watch?v=yx7-yvXf6f8
A lot of research went into everything from the gender to the tone of the voice so that pilots would know to immediately pay attention to the warning system above everything else.
And now and off-topic question: why would rolling right stop a plane from going in the sea?
PDF Warning: Page 121 - Aerodynamic Forces in Flight Maneuvers
https://www.faa.gov/regulations_policies/handbooks_manuals/a...
Any time you bank the wings of a plane to make a turn, you have to also pull back on the stick to maintain the same amount of lift to keep you in level flight. The more you bank, the more 'back stick' you need to maintain the lift.
It's a logarithmic corellation, i.e. in a 90 degree banked turn, it is impossible (you need infinite lift) to maintain altitude.
Most aircraft get to the point where the bank is just to steep and you cannot generate enough lift to hold your altitude any longer. Pulling back on the stick some more will just end up bleeding off airspeed and putting you into a stall/spin. At that point your only recourse is to go back to 'wings level' and rebuild airspeed.
And according to that wikipedia link, most of it was wrong or conflicting!
Those high-pitched piezo speaker beepers are cheap and simple, but it is difficult to localize where the beep is coming from, and the single frequency can't communicate much. Mostly, you can't tell if your gate beeped, or the one next to you.
His solution was, among other things, to turn it from a beep to a bell tone. That alone made it better.
It seems like the audio design of things is often missed. I'm excited about products like Google's Home and Amazon's Echo because we're finally considering audio as a first-class user interface.
Probably not:
In an article in Vanity Fair, William Langewiesche noted that once the angle of attack was so extreme, the system rejected the data as invalid and temporarily stopped the stall warnings. However, "this led to a perverse reversal that lasted nearly to the impact: each time Bonin happened to lower the nose, rendering the angle of attack marginally less severe, the stall warning sounded again—a negative reinforcement that may have locked him into his pattern of pitching up" which increased the angle of attack and thus prevented the plane from getting out of its stall.
Thats like suddenly having your car freeze the speed gauge or something because the speed would be absurd. Who the flip comes up with this things?!
Sleep as Android has a feature where you need to perform a conscious task (math problem / tap on X jumping sheep / etc). Sure you might still solve the math problem and go back to sleep, but there is no doing multi-operational math subconsciously. You're going to know you are awake and will remember.
To turn off the alarm, I'd have to solve 3 problems like "2x5+14-6" without getting any wrong. I eventually got really good at doing math in my sleep apparently, because I overslept a few days after about a year of using it. Later finding that my alarm was in the solved state.
Now I use the mode that makes you physically shake the phone a certain number of times (and a certain amount of force) to turn off the alarm. I have it set so it takes about 20 seconds of vigorous shaking, and by the end of it I'm fairly awake.
I did the ringtone thing for a bit and it just made me really hate my ringtone. I would never hear my alarm and go "maybe someone is calling", but I would always hear my ringtone and go "Wtf, why is my alarm app...? oh..."
Perhaps such environments could benefit from a button that silenced all current alarms for at least 30 seconds, leaving the operator a brief quiet time to focus, modulo new alarms.
Asking the annoying manager to go to some mildly distant place to find an appropriate manual or checklist binder (that may even exist!) may work well for some programmers and some managers.
For a more common example, spend a few minutes in the lobby of a busy or understaffed McDonalds and listen to the cacophony of beeps.
signal-to-noise is hard.
This. It doesn't matter how clever you make the alarm. If it mostly goes off in the absence of real danger, you are teaching people to ignore it.
She was in a car with a bunch of her friends who were also from Seattle. She kept asking what she should do. I kept telling her that they should pull over and seek shelter in a building. Her follow up question each time was "Is there really a tornado? What do we do if there is a tornado?"
It was really strange from my perspective, because I've done so many tornado drills the response is second nature. That being said, if there was a major earthquake I'd probably just freeze for the entire duration, and only remember to hide under a desk or in a door frame after it was over. The first time I felt an earthquake I was 23, and I didn't even realize what happened until I read about it in the news later.
It seems like one of the problems with designing an alarm for a particular level of urgency is that people react in different ways to stress. My ex was not calm under pressure, and would totally freeze up. I'm reasonably cool under pressure, but alarms will derail my train of thought even when I'm concentrating on the issue at hand. I personally think that unless the necessary course of action is extremely trivial, such as "leave the area immediately and seek shelter", constant alarms are more of a hindrance than a help. I like alarms that have a brief alert to get your attention, and then proceed to give a lower-level alert for a while, with the more pressing alert sound playing again as long as the issue is still present.
For instance, one of the the best design features I've found in my Subaru is how they handle traction control. In the Toyota's I've driven, when Traction control comes on the system starts chirping like bearing that's about to fail at 10k rpm. In my Subaru, a light flashes on the dash, just in your peripheral vision. It's definitely helpful to know TCS is being activated, but I would hardly call it an emergency.
The two times TCS seems to be activate are 1) when pulling out from a dead stop, and 2) during emergency maneuvers. In the first case, a constant alert is unnecessary. For one thing, you'll probably notice TCS has activated when it cuts the throttle. For another thing, if you can't immediately stop whatever you are doing, it's probably because you have a more pressing matter, like situation 2. I had a cab cut me off to pull a u-turn from the right last year when the roads had 14 inches of snow on them. I was about 8 car lengths from the cab, and there was a car stopped in the oncoming lane waiting to turn left. I slammed on my brakes, and realized I would never stop in time. I steered to the left and hit the gas while leaning on my horn. The cab slammed on it's brakes and I passed just in front of it. The car on the left side was about 3 car lengths past the cab, and I was already on the brakes and spinning my wheel left to avoid it. I immediately had to turn back to the left because I was pointing towards the opposite side of the intersection. I gave my car a blip of gas to straighten back out.
In this process, my traction control came on three times: once when initially slammed on the brakes (likely tied to ABS activation), once briefly when I turned back to the left to avoid the oncoming car (likely preventing me from sliding in to a frontal offside collision), and once when I tapped the brakes trying to avoid the far corner of the intersection (likely preventing me from running off the road ass-backwards.) At no point would the methodology of the Toyota be helpful in that situation. I needed to focus 100% of my concentration on the road and the two cars, not on the fact that my car was helping me not die.
On a similar note, the fact that Subaru builds a bit of slide in to their traction control is very reassuring. It gives you a lot of feeling for where the limits of the car are. I think I probably would have just ridden out the t-bone had I been in a Camry; I have no idea where the limits are on those cars because the traction control locks down and slide immediately, and the steering has all the feel of a limp-wristed handshake. Of course, a Camry is also FWD, and I had snow tires on my car, so I probably was getting some steering help from torque vectoring in the rear. I was also really grateful that my mom had taken me to empty parking lots when I was learning to drive and had me drive (and slide) around until I understood how cars lost control, and what to do.
It's Federal Signals' "Alternate Wail" setting for their outdoor warning products. Multiple overlapping arrays making this sound is (as you can tell) very effective at making people take shelter.
Perhaps you've had some musical training, so you're more sensitive to this effect?
In comparison, the tornado sirens someone linked to are really noticeable and annoying to me.
The two most effective things that I've done to wake myself are:
a) I had my computer play a song with a high BPM at 100% volume on repeat starting on a timer. Shutting it off required typing a few commands to kill the process. This was effective for a while until I started sleeping through it too.
b) What I do now is I have a cheap clock radio. I set the radio dial to an ultra-conservative talk radio station and then superglued the dial in place. I also superglued the volume dial at its max position and broke off the snooze button. This is highly effective at getting me up and motivated to leave the house.
Works for me, at least. Might be worth a try for you, too. I should think almost anything would be preferable to the self-abuse of loud talk radio.
[1] https://en.wikipedia.org/wiki/Reveille - has a playable sample
I'm a heavy sleeper _and_ a dawdler though. The radio thing has the extra effect of motivating me to get out as quickly as possible.
The real problem for me is wintertime, because then I have to wake up in the dark. That just makes it super hard to get going in the morning.
Anyone hear a weird beeping noise? Wonder what it could mean...oh well, I'm sure it's not important, we're about to land anyway.
https://www.youtube.com/watch?v=5McECUtM8fw
Geez that sound is annoying, I hope it stops soon.
That first landing looked far from 'perfect'. He was coming in high and possibly too fast, and I daresay he was focusing too much on how much runway he would have left after landing. He could have also assumed that the warning was the overspeed warning from extending landing flaps at too high a speed (though most light aircraft don't have that warning system).
That second one was a lot less explicable - looked like a normal landing approach. Surprised that the instructor also didn't pick it up.
Interestingly, that second pancake landing is in exactly the same type of aircraft that I did my initial flying training in. Something about seeing that instrument panel, and hearing that warning alarm made the hairs on the back of my neck stand up, and raised my pulse a little. Even 25+ years after I last flew that type!
The problem occurs when you choose a phone alarm sound that is frequently used in the industry for seriously nasty things. Like frequently used in many movies for the biggest disaster that will be occurring, leaving you in a panic to turn off the sound on your phone right in the middle of the most important part of the movie.
Or, you work in the central IT Services department of a major university, and you work in the building where the main computer room is located. And your alarm sound happens to be the same one as used on the floor for the halon dump fire suppression system. And you're standing at the doorway of a co-worker talking to them, when your phone alarm goes off. And their office happens to be right next to that of the datacenter manager.
So far as I know, there were no heart attacks or strokes that resulted from my accidental use of the same alarm library sound that day. But I was strongly incentivized to switch my alarm sound to something else.
I'm not an alarm designer, but my understanding is that if you want pretty much everybody to be able to hear something, not just your kids and your dog, 10-12KHz is a much better upper limit.
I found an Android alarm app that tried to wake you when you're in a shallow sleep phase. IIRC, the app would quietly ring 45, 30, and 15 minutes before the wake time you set, so if you happened to be in a shallower sleep phase then you would wake more naturally. I stopped using it, though, because the phone UI was too annoying to use as an alarm clock.
The "bacon alarm clock" is also a real thing, being basically a timer-driven toaster oven with precooked bacon in it, but that'd be a bit more worrying to run automatically than just a coffee maker.
A few months later we understood. The first few times you see a bellowing metal beast with its horn blaring, you can't help but to watch out for it. A few hundred safe encounters later and it actually becomes very easy to ignore it. There were some other near misses that year.
Personally being woken up to something funny, cute, meaningful from a friend or loved one was great. Though the challenge for this idea to work is you have to put restraints via audio volume levels to ensure a great experience (waking up laughing or with a smile or another positive emotion) happens on each use. Also know the likes of each user and what's happening in their life.
I still believe it's a great idea/new form of communication that if done correctly the masses would love, as when it wakes you up to something you connect with you smile and or laugh; changing your morning routine for the better! I wish Apple would allow a user the ability to wake up to a radio URL!!!
As a heavy sleeper I'm always groggy and not social when I get up. My perfect alarm clock would have a loud, blaring alarm (imagine what a movie alarm would sound like to indicate nuclear war or some other form of certain doom). I think I'd either be annoyed by or sleep through something cute like laughter.
I keep thinking I should get back to working on it and finish it the way I described above...
Not sure it's even right to call it a squeak. Many adults hear nothing at 20,000 Hz.
In airliner cockpits this also became a problem when CRTs superceded analogue instruments as the constant drone of their cooling fans masked important aural cues. It has improves with the use of cooler LCDs but even now a 737 cockpit is shockingly noisy.
ts
What's more, the beeping happens at the patient's bedside. Of all the actors who are in a clinical hospital environment, it is the patient who is the absolute least empowered to do anything to change the situation, but who is also the most negatively impacted by an alarm. Loss of sleep is a major problem for hospitalized patients[1].
Perhaps nurses respond more to patient beeps/complaints than to the alarms themselves, but it's a bit sadistic if you think about it (though unintentionally).
[1] http://journals.rcni.com/doi/pdfplus/10.7748/ns.29.28.35.e89...
I might have to start checking Atlas Obscura more often if they have articles like this on it.
That's because they are CYA lawsuit deflectors.