Slashdot Mirror
System Measures Stress In Emergency Callers' Voice
cylonlover writes "Chances are that if you're calling 9-1-1 (or 9-9-9, or whatever it is where you are), you're not likely to tell the operator that your case isn't all that urgent, and that it can wait. The problem is, sometimes emergency dispatch centers are so overloaded with callers – all of them stating that they need assistance right now – that some sort of system is required in order to determine who should get help first. Dutch researchers claim to have developed just such a system, which analyzes callers' voices to determine how stressed-out they are."
Afghani freedom fighters organise a denial-of-service attack by playing back Frodo in the Lord of the Rings to the telephone.
How well does this thing work with child callers, or those with developmental disabilities who do not respond 'normally' to emergency situations?
I had a girlfriend who's mom would freak out at the most silly things, and not so silly too, accentuating her voice to make this overtly apparent.
Should have seen her when I accidentally ran over her cat. Very unfortunate, and people react very different in panicked, or life threatening, situations.
I wonder how well this detection will hold up, 4% margin of error seems quite low.
This is very nice from the signal analysis perspective, but the implication that emergency call may be delayed if the caller is not stressed is a bad idea
How well people deal with emergency situations varies hugely. This system would prioritise a 5 year old ringing about a huge splinter she just got over a military veteran reporting a 3 car pileup with limbs everywhere. Can't beat human judgement in a job as important as this.
However, the stress in the callers voice may not indicate the severity of the emergency.
Some people can be calm and collective in very high stress situations, whereas some people freak out when someone has a dizzy spell. Additionally an outside observer may be less stressed, for example someone calling in a 5 car pile up or reporting that someone in their store just collapsed. And then there are children making calls... which probably introduced a whole new level of random.
The article mentions an error rate, but doesn’t really seem to elaborate as to whether that error rate is stress to emergency, or the algorithm’s ability to identify stress. Before deploying something like this, I hope they do some kind of study to determine if stressed voices correlate to actual emergency severity in the majority of cases (which they may have already done, the article isn’t clear).
how it will deal with a Scottish accent .
Caller stress doesn't correlate well with how important the call is. It correlates with how closely involved the caller is in the incident.
Besides a lot of people will panic like crazy at, say, a small car accident where no-one was hurt.
This may seem counter intuitive, but it's a horrible idea. This will provide artificial priority for the histrionic personality type.
I have seen people get hysterical over a fender bender screaming "Oh my baby!" and I have seen people laughing while trying to control a broken airplane. Just hire dispatchers with a bit of common sense.
This is like the TSA always trying to find a machine to do the job that a human could do way better if they were allowed to do it with common sense.
So people that remain calm and do not panick during an emergency will get lower priority and have to wait, whereas people that totally freak out and start to cry because the cat doesn't come down from the tree will get help immediately.
Should emergency dispatch centers be staffed by enough people that are adequately paid instead?
Natural selection will provide some interesting long-term consequences.
Can we get a "-1 Wrong" moderation option?
Haven't you heard? Emergency services has a new number. It's 0118-999-881-999-119-7253.
SmashTech - No smashing of tech involved
I'm getting tired of this trope, especially as it's used as keyword to get immediate panic reflex.
The fact is, in any given society, resources for healthcare will be limited, and generally smaller than what is perceived as necessary by the public.
This will automatically cause a need for prioritizing, as some medical threats are more immediate than others, and should be treated first.
These researchers have been looking into a way for making that more effective. However, it has not been implemented! So 'the Dutch' aren't doing anything here.
Given that waiting lists have been exceedingly long in NL for quite some time now, not due to lack of funding per se, but lack of trained personnel, it's also more than a little irrelevant.
Lastly, I'm just going to assume you live in the U.S.A here, since you're using the rationing healthcare rhetoric. May I remind you that this is done on a large scale in your country already? Only in your case, it isn't rationed based on need, as any decent person would want, but based on how much money you have. Yes you can, in a few select places in the U.S.A, get the best possible healthcare, but only if you have the enormous amounts of money that's asked for it. Normal people have to do with less healthcare than any given Dutch person gets, for much more. Rationing is not so much our problem, as yours.
I don't understand this article at all because emergency dispatching is not prioritized based on the caller's choice of priority. I could have ten calls at once all insisting they are the top priority and that information would be irrelevant. The nature of the emergency is what's important, not how badly the caller wants assistance.
I dispatched during the L.A. riots and believe me every caller wanted someone to help them RIGHT NOW and I don't blame them. But calls for people being beaten got priority over property crime calls. I question the thought process behind this article that dispatchers do not or cannot already properly prioritize calls.
It does not take into account 3rd party callers either. For example, I'm driving along and see a horrible accident right in front of me.
I call in, calmly describe the detached body part in the road, the 5 cars involved, and the blood on the pavement. Then compare it to the guy who lost the blood or the finger or whatever calling in completely hysterical.
Same situation, one person much calmer than the other. It is a common occurrence. Ditto friend vs spouse or parent or child.
Definitely cool technology, but assuming the article is accurate, not a good usage.
Is it just me or does it seem like keeping a lot of people on hold when they call 9-1-1 is a bad idea? Seems to me the better approach would be 1. Stiffer penalties for non-emergency callers 2. More people answering calls You could even have lesser trained staff to handle the less serious calls in order to keep lines open to the fully trained staff and defray cost.