Real-Time Data from Amsterdam’s Fire Service

Posted on 01/31/2012 by


Brandweer EngineBart van Leeuwen has been well-known in the world of Linked Data for the past few years, sometimes called “the Semantic Fireman.” His story is easier to remember than many other ambitious discussions of using data to improve effeciencies, I am guessing because of the way it can begin:

It is your job to guide the driver to the location of the incident, your only help are old sun-bleached papers with driving directions typed out in a 10 point font. Since you are new on the job, there is no room for error.
Nodalities Magazine, Issue 12

He has been putting data to work to improve the fire service in Amsterdam, and I was able to catch up with him recently about the data he’s now publishing in Kasabi. The dataset is being continuously updated, so the information is essentially real-time. This makes it even more interesting, because Kasabi is being put to work as a place to publish and find constantly-improved information.

Bart’s dataset covers all incidents where units of the fire department (Brandweer Amsterdam Amstelland) are sent, including geolocation and street names—not full addresses. It’s made up of events, places, and some department-specific attributes. You can see examples of the data here. Places are encoded in VCard and geo, with a full list of vocabularies and classes available from the schema page.

I asked Bart what the idea behind the dataset was:

Open government. The fire department wants data from other departments, so why not publish data ourselves? Furthermore, it is possible to make us more visible to the public.

Where does the data come from, and how is it updated?

Currently, it’s agregated from the pager network. But, while writting this, we are preparing to get it straight from the dispatch systems. The data is realtime, if a fire engine leaves the station, you can find where it goes on the web of data.

Who do you see using the data in the future?

There is not that much use yet, but potential users are: news sites, bloggers, statistical institutions but also app developers.

Why are you using Kasabi to publish the data?

We use real-time Linked Data internaly, and I didn’t feel like putting our daily operations under presure by linked data users. Kasabi provides a nice platform to upload the data, and provide some nice API’s for it. We can also see the usage statistics for the data, and look forward to doing more with that in the future.

Finally, what is the future of this dataset: where do you see it going in terms of updates or improvements?

With the attachment to the dispatch system, we will have more fine-grained information and better relations between the various dispatch messages. Currently, we can’t say for sure if one unit is going to a previosuly used address if the incident jut got bigger. Another very important update is the usage of external links— e.g. geonames, dbpedia and some Dutch government datasets describing legitslitve areas. Besides that, we are building and promoting to hold a library of firedepartment related terms, we will surely link that in the futur as well.

I’d like to thank Bart for taking the time to talk with me, and encourage you to have a look at the Brandweer Amsterdam-Amstelland dataset. Let me know what you think, and if you have any questions or ideas on how to use this kind of data. Either post a comment below, or drop me an email:

Dataset mentioned:

Image: “Blaricum Fire Department” by Floris M. Oosterveld via flickr CC: by 2.0

Posted in: Datasets