Monte Carlo firefighting

I know I’m making a bit of a gamble here with the title. After all, the category is rescue and the gist of the title comes from Computer Science, algorithm design and analysis to be more precise. The two fields couldn’t be much further apart…

FireHouse.com had an article yesterday on how small towns that rely on volunteers or oncall firefighters have problems in getting enough personnel to respond to calls, especially during the daytime. It is nice to notice that the same problems plague the States as Finland in this respect.

My background is that of a volunteer firefighter turned into a oncall firefighter (part-time) and finally into a systems designer/programmer/web developer/IT-guru at the local rescue (fire) department. For over a year of working full time on the IT systems at the main station, which is fully manned 24 hours a day, and participating to calls at my own unit outside of business hours, I finally got transferred to the my own unit to work during the days. Now at least the unit has one person responding to calls during business hours. And we are the first responders in a town of almost 5000 inhabitants.

In North Karelia we have 19 counties, of which most rely on either volunteer or oncall firefighters to respond to calls. Some of the counties have one or more people on duty during the daytimes, but generally the amount of paid personnel isn’t enough to handle even a small house fire or traffic accident. The problem that small towns in Finand and the U.S. face is twofold:

  • Most of the volunteer and oncall firefighters have day-jobs that prevent them from responding to calls during office hours. This is even more pronounced in towns that are mainly bedroom communities.
  • The amount of volunteer and oncall firefighters is alarmingly low and getting new personnel is difficult.

The FireHouse article describes the situation quite well. In those towns that aren’t bedroom communities, firefighters can always negotiate with their employers so that they could respond to calls. At least if the call is to something big enough. But especially in bedroom communities this approach can’t offer any respite. The town in which we live is exactly such a bedroom community. Of all of us oncall firefighters, I’m currently the only one working or studying in town. Hiring personnel to serve at these stations during office hours is prohibitevly expensive. Our rescue department has personnel costs of about 80% of the yearly budjet. Hiring one to four people per station really isn’t an option.

Why one to four you ask? The Finnish Ministry of Interior (the top officials in fire and rescue in Finland) has a guideline that defines a rescue unit that consists of an officer, driver and one to three pairs of firefighters. So the minimum size of a unit is four people, the optimum size is six. The truth is, our town has ten firefighters on the roster, of which eight are currently active. At any given time, it is quite uncertain that out of the eight at least 50% would be available to respond to calls. Having a larger pool of firefighters available would help the situation, but the relief wouldn’t be immediate even if we did get an influx of new recruits.

Being a volunteer or oncall firefighter isn’t exactly an easy hobby. The physical requirements are quite tough and you have to be ready to drop everything you are doing and rush to the station house when a call comes. It isn’t something that is easy on family life. The training required to do the work is quite lengthy, and even then it isn’t anywhere near being over. Life-long learning is very much a fact for all firefighters, be they volunteers or professionals. How to get new and interested recruits is still something we are very much trying to figure out. Let alone how we are going to find the time to train them.

And what did Monte Carlo have to do with anything? Well, the original Firehouse article had dice rolling in the title. After all it is a gamble to see how many will respond to a call. Monte Carlo algorithms are a group of solutions to problems that do not have computable results in a time that would be meaningful. They use sophisticated guesswork to try to achieve a good enough result. Sometime the gamble pays off, sometimes it doesn’t. (This is an overly simplified explanation, but it should do.)

There aren’t any easy answers to the situation. It was just pleasing to know that we (Finns) are not the only ones with the problem.

P.S. The Secret Service seems to have its own fire department for the White House and VP’s residence…

Leave a Reply