NYC: Big fire (4th alarm now) across a church and several multi-unit residential buildings at 12th St and 27th Ave in Astoria.
-
NYC: Big fire (4th alarm now) across a church and several multi-unit residential buildings at 12th St and 27th Ave in Astoria.
5th alarm now w/2 extra tower ladders.
Queens Box 7509 for those of you with a tapper circuit at home.
-
NYC: Big fire (4th alarm now) across a church and several multi-unit residential buildings at 12th St and 27th Ave in Astoria.
Several firefighters injured after a collapse in the church. Now a fifth alarm (the maximum pre-determined response).
The FDNY "alarm" system addresses an interesting and difficult optimization problem in dispatching additional firefighters to large incidents. And it's largely unchanged from the 19th century. It's really quite clever.
Some background:
-
Several firefighters injured after a collapse in the church. Now a fifth alarm (the maximum pre-determined response).
The FDNY "alarm" system addresses an interesting and difficult optimization problem in dispatching additional firefighters to large incidents. And it's largely unchanged from the 19th century. It's really quite clever.
Some background:
@mattblaze ooh, good, I've wondered about this term and now I will learn more
-
Several firefighters injured after a collapse in the church. Now a fifth alarm (the maximum pre-determined response).
The FDNY "alarm" system addresses an interesting and difficult optimization problem in dispatching additional firefighters to large incidents. And it's largely unchanged from the 19th century. It's really quite clever.
Some background:
The FDNY assigns every location in the city to its nearest fire alarm box, which are (or were) every block or two apart throughout the city. If someone pulled the alarm box, the nearest available firehouse would respond.
(Until surprisingly recently this was done with a spring-loaded telegraph encoded in the box that would ring bells in a particular pattern in each firehouse identifying the box and which engines and trucks should respond).
2/
-
The FDNY assigns every location in the city to its nearest fire alarm box, which are (or were) every block or two apart throughout the city. If someone pulled the alarm box, the nearest available firehouse would respond.
(Until surprisingly recently this was done with a spring-loaded telegraph encoded in the box that would ring bells in a particular pattern in each firehouse identifying the box and which engines and trucks should respond).
2/
The box technology has changed (most surviving ones can do voice communication now), and many of the physical alarm boxes have disappeared, but every location is still associated with a box, which, even if the box isn't there, determines which firehouses are "first due".
When you call 911 today, the fire dispatcher first figures out the box number associated with the location and sends the first due engines and trucks for that box.
3/
-
5th alarm now w/2 extra tower ladders.
Queens Box 7509 for those of you with a tapper circuit at home.
-
The box technology has changed (most surviving ones can do voice communication now), and many of the physical alarm boxes have disappeared, but every location is still associated with a box, which, even if the box isn't there, determines which firehouses are "first due".
When you call 911 today, the fire dispatcher first figures out the box number associated with the location and sends the first due engines and trucks for that box.
3/
The first due units are pre-determined, but are almost always the nearest firehouses, which makes sense to minimize response time.
But what happens when the fire is large enough to require more firefighters than the first due?
The fire chief on the scene can call for additional units (originally by tapping a telegraph key inside the box), but who should come?
You might think it should be the NEXT nearest firehouses, expanding outward as more and more are required. But that's a bad idea.
4/
-
The first due units are pre-determined, but are almost always the nearest firehouses, which makes sense to minimize response time.
But what happens when the fire is large enough to require more firefighters than the first due?
The fire chief on the scene can call for additional units (originally by tapping a telegraph key inside the box), but who should come?
You might think it should be the NEXT nearest firehouses, expanding outward as more and more are required. But that's a bad idea.
4/
@mattblaze is it bad because it drains the general ability to respond to other issues in the area?
-
The first due units are pre-determined, but are almost always the nearest firehouses, which makes sense to minimize response time.
But what happens when the fire is large enough to require more firefighters than the first due?
The fire chief on the scene can call for additional units (originally by tapping a telegraph key inside the box), but who should come?
You might think it should be the NEXT nearest firehouses, expanding outward as more and more are required. But that's a bad idea.
4/
The problem with always dispatching the nearest available fire house is that if there's a large fire somewhere, there won't be nearby available firefighters to respond to *other* fires nearby, requiring the initial response to come from far away.
So they don't always send the nearest available. Instead, they skip over some of them, to ensure maintaining availability near large incidents.
But figuring who best to send next quickly becomes pretty complicated, with multiple contingencies.
5/
-
The problem with always dispatching the nearest available fire house is that if there's a large fire somewhere, there won't be nearby available firefighters to respond to *other* fires nearby, requiring the initial response to come from far away.
So they don't always send the nearest available. Instead, they skip over some of them, to ensure maintaining availability near large incidents.
But figuring who best to send next quickly becomes pretty complicated, with multiple contingencies.
5/
So, instead of requiring the dispatcher to figure this out on the fly (which would be impossibly difficult to do well when multiple fires are going on, especially before computers), they figure out the response order for each box in advance.
Each box has a "box card" listing the first due, second alarm, third alarm, and so on responses. Each alarm increment is a contingent of several engine and truck companies. The box card is used to decide who to send when additional alarms are called.
6/
-
The problem with always dispatching the nearest available fire house is that if there's a large fire somewhere, there won't be nearby available firefighters to respond to *other* fires nearby, requiring the initial response to come from far away.
So they don't always send the nearest available. Instead, they skip over some of them, to ensure maintaining availability near large incidents.
But figuring who best to send next quickly becomes pretty complicated, with multiple contingencies.
5/
@mattblaze this feels like there's some kind of maths optimization problem here with a theorem to be proved
-
So, instead of requiring the dispatcher to figure this out on the fly (which would be impossibly difficult to do well when multiple fires are going on, especially before computers), they figure out the response order for each box in advance.
Each box has a "box card" listing the first due, second alarm, third alarm, and so on responses. Each alarm increment is a contingent of several engine and truck companies. The box card is used to decide who to send when additional alarms are called.
6/
The box card ordering ensures that there remains some available coverage spread throughout the city when there are large incidents going on. They pre-calculate up to a fifth alarm (at which point there are over 35 engines and trucks operating at the incident). After that, more units can be called in, but the dispatchers have to figure it out on their own.
(9/11 was a fifth alarm, but had many more fire companies than that called in).
7/
-
The box card ordering ensures that there remains some available coverage spread throughout the city when there are large incidents going on. They pre-calculate up to a fifth alarm (at which point there are over 35 engines and trucks operating at the incident). After that, more units can be called in, but the dispatchers have to figure it out on their own.
(9/11 was a fifth alarm, but had many more fire companies than that called in).
7/
Anyway, it's a very interesting and complex real-time systems optimization problem, with lives at stake. And it was solved with 19th century technology, with the basic principles unchanged to this day.
8/8
-
Anyway, it's a very interesting and complex real-time systems optimization problem, with lives at stake. And it was solved with 19th century technology, with the basic principles unchanged to this day.
8/8
@mattblaze Thanks!
-
Anyway, it's a very interesting and complex real-time systems optimization problem, with lives at stake. And it was solved with 19th century technology, with the basic principles unchanged to this day.
8/8
typo "unhanged" should be "unchanged"
-
Anyway, it's a very interesting and complex real-time systems optimization problem, with lives at stake. And it was solved with 19th century technology, with the basic principles unchanged to this day.
8/8
@mattblaze Fascinating, didn't know any of that. Do you know if it is an approach specific to NYC, or something large cities tend to adopt?
-
@mattblaze Fascinating, didn't know any of that. Do you know if it is an approach specific to NYC, or something large cities tend to adopt?
@tehstu NYC has always been on the leading edge of this, but other dense cities generally do something similar.
-
Anyway, it's a very interesting and complex real-time systems optimization problem, with lives at stake. And it was solved with 19th century technology, with the basic principles unchanged to this day.
8/8
@mattblaze small clarification aside: I assume 'companies' here is a term meaning something similar to 'crew' or 'team', rather than implying some kind of privatised outsourced structure?
-
@mattblaze small clarification aside: I assume 'companies' here is a term meaning something similar to 'crew' or 'team', rather than implying some kind of privatised outsourced structure?
@uep Today yes, but it has its roots from the time when there wasn't a single unified fire department in the city.
-
The box technology has changed (most surviving ones can do voice communication now), and many of the physical alarm boxes have disappeared, but every location is still associated with a box, which, even if the box isn't there, determines which firehouses are "first due".
When you call 911 today, the fire dispatcher first figures out the box number associated with the location and sends the first due engines and trucks for that box.
3/
@mattblaze Boston's fire boxes are very much still around, and although the backend systems have been computerized for quite a while now, the boxes themselves are still very much a wind-up telegraph inside. As a nice feature, the wiring is entirely independent of the phone and electrical systems, so should there be a major systems outage in one area, the boxes still work. They actually did work last time 911 went down state-wide and someone pulled one out of desperation. Thankfully the news reported on that, and it was promoted enough for people to actually read it and learn those are still actively maintained and monitored...
The fire museum has one of their oldest computers, as well as a disconnected box for people to pull and look inside of. (Well, it's connected to the computer behind it, but not to the actual fire system)
And yes, they still have the manual telegraph key in the box. AFAIK it's only really used for testing stuff these days?
