Fighting fires with data

How data can give emergency services an edge

With far greater government attention to field data processing and communications, fire and emergency services can better handle some of the more bizarre and dangerous situations they encounter.

This is the message of Netherlands-based fire data consultancy Netage, run by Bart van Leeuwen, a senior firefighter with the City of Amsterdam Fire Department.

Australian government agencies are mulling over some of the Van Leeuwen prescriptions following a presentation earlier this year at AFAC17, the annual conference of the Australasian Fire and Emergency Service Authorities Council.

The invitation for Van Leeuwen to present came from Australian emergency specialists who heard him at similar conferences in North America earlier in the year.

Netage has been operating from a village close to Utrecht for the last decade, taking its message of linked data and smart emergency services to clients around the world.

Two themes of the Netage computing doctrines are giving operational fire crews quick data on the run for critical, unusual events — outliers, in statistical parlance — and developing a common definitional language for firefighting terms, for use by data preparation experts.

As a new recruit to firefighting in the Netherlands, Van Leeuwen was told by his assigned mentor that the department never received normal calls.

Take, for example, a call to a residence where the occupants were waiting for an interstellar mothership to transport them to another star system, and didn’t seem to share the arriving firefighters’ preoccupation with a blaze flaring up in the kitchen in the here and now.

After that, it probably seemed relatively mundane to attend an emergency at a home where the owner had a collection of more than 400 vacuum cleaners.

When it comes to succinct but accurate and informative data quickly processed and transmitted to field crews, how do you characterise a plane crash on a railway line just short of a road underpass? Is it a fire on a railway, a plane crash, an accident near a bridge over a road?

A voice operator might handle it in a few pithy words, but an automated data processing system needs to have some carefully thought-out protocols to present all that clearly.

As a Dutch operation selling to more than a few agencies in English-language jurisdictions, the Netage consultancy is keen on winnowing out a lot of conflicting firefighting terminology.

For example, just among the fire specialist audience at one talk, scalers for some were roof ladders for others; some drove water tankers but others had water tenders; some would hang a plug where others would hit the hydrant; and what was a hood tool for one lot was a ray tool for another (named after the person who introduced its use).

After several years’ work, the Dutch have assembled a “Firebrary”, a standard dictionary of agreed firefighting terminology.

The concept of linked data extends to the Internet of Things; that is, the developing definitions and protocols facilitating routine transfers of data between different items of machinery, equipment and installations.

New Amsterdam fire trucks are “open data fire engines” that report how much water they are using, how full their tanks are, how much diesel fuel is left, are there toxic fumes in the air, in what concentrations and other useful field data.

In Australia, recent improvements in government data processing, information handling and telecommunications systems include satellite-based position-finding and timing, and rationalisation of disparate computer systems.

Familiar big names such as IBM, SAP, Motorola, Oracle have handled massive hardware investments, and migration of huge banks of data. The Dutch fire data system is a complementary, cloud-based concept, a definitional and software-oriented offering designed to be added to and work with other suppliers’ basic infrastructure.

The sort of data firefighters could use in a hurry covers a wide range of database systems, infrastructure, software, communications protocols, information structure and definitions. It could be map databases, housing records, building usage limits, storage whereabouts of hazchem items, layouts of real estate locations, ownership records, and details of temporary road closures.

Various semantic web technologies are integral to most of the output of Van Leeuwin’s firefighting data work.

“It’s huge volumes of data, hundreds of gigabytes,” Van Leeuwen explained during an interview in Sydney. The company has never invested in hardware, instead minimising capital costs by relying on cloud-based infrastructure.

“It offloads the maintenance burden,” he said. “Even if potential new clients want a test, you only have to hire time for the actual test.”

“The reason I stick to fire is that there are very few people that understand the fire industry,” Van Leeuwen. “The innovation comes from introducing them to computing. We apply existing technologies.”

He said his message had found a receptive audience in Australia.

“I have a feeling I’ll be back at some date,” he said. “The good thing is that all the people I talk to say it needs to be done, making better use of data. We could use data for community risk reduction.”

He argues this still requires some evolution in the mindsets of working firefighters and emergency personnel, not just in Australia, but in North America and Europe as well.

“It is time for the fire services to include learning about computer issues and data in their training,” he argued.

“In the US you need either a Masters or a PhD for higher executive levels in firefighting, but there’s nothing on data science.

“It’s not an accepted part of the curriculum in Europe. People are not listening yet, but they will.”