Author(s): Casey Grant. Published on March 1, 2018.

Eve of Revolution

How the use of predictive analytics to improve ITM processes suggests the next data-driven chapter for safety professionals

Clive Humby, a data scientist from the United Kingdom, is widely credited for being the first person to say that data has replaced oil as the world’s most important commodity.

Because I’m curious, I checked the validity of this analogy by reviewing the market capitalization of the world’s largest public companies over the last six years. I found that, in 2011, six of the top 10 largest companies were in the petroleum business, while only two were technology based. By 2017 this had flipped—six technology giants were now in the top 10 in the world in market capitalization, while only one oil company remained on the list. Humby was right: figuratively and literally, data is indeed the new oil.

Much like the rest of the world, NFPA and the public safety industry have embraced this paradigm shift and are working diligently to harness data’s immense and transformative power. As a result, I believe we are on the cusp of a data revolution that will improve how NFPA and safety professionals make decisions, optimize our reliability, and maximize our efficiency and effectiveness.

A great example of this is a current Fire Protection Research Foundation (FPRF) research project using data analytics to answer some lingering questions about inspection, testing, and maintenance (ITM) for fire pumps. Good ITM is critical for the continued proper operation of fire pumps, sprinklers, and other fire protection equipment. Although the systems might not be needed for years, when a fire does start they must function immediately and as designed. Failure is not an option.

However, the NFPA standards on the particulars for how often to test and work on fire protection systems to ensure dependability have been hotly debated for years. The crux of the debate seems to always come down to “show me the data.” Currently, there isn’t much scientific basis for why the ITM frequencies are what they are in the standards.

This is where the FPRF project aims to help. We have collected fire pump data from facility operators, inspectors, insurers, and others, which is an important first step. It’s one thing to have the data, but it’s quite another to make something genuinely worthwhile from it—and that’s where data analytics is key. By comparing and analyzing historical data, computer algorithms can then make predictions about future outcomes. We want and need this for the world of fire safety.

Using predictive analytics on our collective fire pump data can clarify elusive aspects of the ITM discussion, such as identifying the greatest failure concerns or outlining occupancy or application dependencies. Perhaps we’ll find in some cases that increasing test frequencies is appropriate due to an identified risk, or we may discover that we need to reduce how often we test equipment because excessive testing may actually be wearing out components. This effort, which is scheduled for completion by the late summer of 2018, will inform NFPA’s ITM standards and is intended to eventually be extrapolated to all types of fixed fire protection systems that depend on ITM, including fire alarm systems, water-based suppression systems, and special suppression systems.

As NFPA continues its transition to an information and knowledge organization, we consider this activity to be a small yet important step as part of a much longer journey. We are anxious to turn on the valves and let the data flow. With a refinery process operating on the power of predictive analytics, our world of tomorrow will have the best possible fire protection solutions.

CASEY GRANT is executive director of the Fire Protection Research Foundation.