Big Data in the Fire Service

By Will Linger

Managing data in the fire service is complicated. Around the United States, almost every fire department retains data about the services it provides. Many departments treat these databases as junk drawers of information and are unwilling or unable to appropriately excavate and interpret the data for their benefit. In Managing Information in the Public Sector, Jay White says, “First and foremost, information is a vital organizational resource. It is just as important as people and money …. Information [technology] has the potential to radically transform the way public agencies do their business.”1 So how does this affect our fire departments?

Tom McEwen and Catherine Miller emphasize the importance of managing big data in the fire service:

“Every fire department is responsible for managing its operations in such a way that firefighters can do the most effective job of fire control and fire prevention in the safest way possible. Effective performance requires careful planning, which can take place only if accurate information about fires and other incidents is available. Patterns that emerge from the analysis of incident data can help departments focus on current problems, predict future problems in their communities, and measure their programs’ successes.”2

In 2011, the National Institute of Science and Technology (NIST) joined fire service executives, scientists, and other professionals at a data summit to discuss the reality of big-data management in the fire service.3 Issues considered included having a common data collection database and language and promoting scientific-based decision making. (3) The summit’s goal was to “develop recommendations for data collection processes as well as a list of data elements that can be consistently collected and analyzed by local fire departments in an effort to measure the department’s availability, capability, and operational effectiveness.” (3)

These are especially important issues for the local department to address. Once local departments begin to tackle their own database issues, the fire service can compile locally retained data in such a way that it would further benefit scientific research. However, the obstacles to overcoming the gaps in our own data collection are common and widespread.

We all grapple with outdated or nonexistent technology. In the public sector we seem to be behind almost every technology curve, often because of lean economic provisions. The summit noted the following four main challenges facing fire departments in managing big data:

  • Burden. The effort required to complete data entry.
  • Usefulness. Recognizing the inherent value data elements have to fire departments and local communities.
  • Motivation. The commitment of the participants to produce accurate and complete reports.
  • Accuracy. The elements need to be measurable and reproducible, and the analysis must yield reliable and effective conclusions. (3)

Despite the obstacles, remember that the information that we collect in the fire service is some of the most important in regard to the safety and well-being of our citizens, so it’s important to do everything to ensure that we are retaining “clean” data. Moreover, the accessibility of this information is vital as we look to improve key benchmarking goals relating to safe staffing levels, efficient deployment models, and the prevention of fire-related death and injury. In recent years, this information has led to valuable research that has challenged some long-held assumptions held by fire service stakeholders.

NIST completed a widely known study, “Residential Fireground Field Experiments,” in which it analyzed the 22 most commonly performed tasks on the residential fireground. The aim was to quantify task efficiency with alternate staffing models.

“The experiments allowed the same number of firefighters, arriving at the same time to the same type of fire, to be compared to the outcome when only one of those variables were changed (crew size). This maximized [the] understanding of the relative effects of response time and crew size to various fire severities and provides validation and insight into the results ….” (3)

“For overall scene time (time to complete all 22 fireground tasks), the four-person crews operating on a low-hazard structure fire completed all the tasks on the fireground (on average) seven minutes faster-nearly 30%-than the two-person crews. The four-person crews completed the same number of fireground tasks (on average) 5.1 minutes faster-nearly 25%-than the three-person crews. On the low-hazard residential structure fire, adding a fifth person to the crews did not decrease overall fireground task times. However, it should be noted that the benefit of five-person crews has been documented in other evaluations to be significant for medium- and high-hazard structures, particularly in urban settings, and is recognized in industry standards.

“There was a nearly 10% difference in the “water on fire” time between the two- and three-person crews and an additional 5% difference in the “water on fire” time between the three- and four-person crews (i.e., 15% difference between the four- and two-person crews). There was an additional 5% difference in the “water on fire'” time between the four- and five-person crews (i.e., 20% difference between the five- and two-person crews).

“The four-person crews operating on a low-hazard structure fire completed laddering and ventilation (for life safety and rescue) 30% faster than the two-person crews and 25% faster than the three-person crews. The three-person crews started and completed a primary search and rescue 25% faster than the two-person crews. The four- and five-person crews started and completed a primary search 6% faster than the three-person crews and 30% faster than the two-person crews. A 10% difference was equivalent to just over one minute.”4

Most recently NIST, in conjunction with Underwriters Laboratories and the Fire Department of New York, conducted a series of experiments on Governor’s Island, New York, to provide quantitative research on, among other items, fire behavior and fire dynamics. This work is important, as we better understand the lighter and tighter modern construction trends and the products used to build and furnish these structures. The experiments were intended to improve firefighter safety and building occupant survivability by examining the following:

-control of ventilation and flow paths to reduce the occurrence of ventilation-induced flashover,
-use of exterior fire attack to reduce firefighter exposures to high thermal conditions, and
-tenability of conditions for victims as a result of these tactics.5

The significance of these studies lies in the scientific data produced, which should influence local fire officers’ decision making. This data analysis also helps us inform higher-level decision makers about why we ask for the resources we do. Our fight to maintain a safe and effective response force is critically important for all stakeholders. Overall, it’s much better if we can make these arguments from a logical platform.

It’s not that these were new issues; but for the first time, it was quantifiably proved that more personnel on the residential fireground makes accomplishing common tasks more efficient. We can also say with some measurable confidence that water doesn’t push fire. Additionally, we can see the real results of door control in a ventilation-limited fire environment. Most importantly, we can maintain a tactically aggressive posture on the fireground because we understand our enemy (fire) better.

Experiments like these marry the power of observation and technology through research. Historically, observation is a key tenet of fire service mentorship. Senior firefighters and officers pass on their experience-based knowledge to younger firefighters. Through experience a senior firefighter may know, for example, that you can cool interior conditions by putting water on smoke when there is no visible fire. But experience is not scientific; it’s a combination of past actions, observation, and intuition. It’s one thing to intuit that our fireground risks are changing and becoming more dangerous; it’s another reality when we can prove it scientifically. Hence, we should understand as much as possible about our potential fire incidents; the research data provided in these reports should, at the very least, help in the vetting of our tactics. Doing so allows scientific research to enhance firefighter safety and increase the survivability of our citizens.

The work of NIST and others is highly valuable considering that it has affected fire service decision making nationally. Note that this information was made public and widely shared. Conducting research, accurate data analysis, and disseminating the information further benefit current and future firefighters. Again McEwen and Miller explain that in relation to data analysis, “Our ultimate objective is to make better and more informed decisions in fire departments. Data (have) no utility in a vacuum, and fire reports stay as data if we do nothing. Analysis turns data into information.” (2)

There are substantial benefits to the task of data analysis for the local department. First, if we manage archived data correctly, we can use this information to assist fire officers in their short- and long-term decision making. With correct data interpretation, fire officers have valuable information with which they can argue for additional staffing, hone standards of cover, and improve firefighter training. (2) This translates into a higher probability of lives saved and property protected. Second, using national standards and benchmarks as the backdrop to local data analysis, we can begin to increase overall professionalism in our departments. The net result is becoming a data-driven organization with the goal of ensuring that our cities are safer overall. Being data driven means to be information driven. And if history is really the best predictor of the future events, then we must pay attention to the past.

In our department, we began our data analysis efforts by examining our training program, because the size and scope of our fire department’s responsibilities have grown. We knew our training program needed improvement. We looked at several key operations statistics, such as historical call type and frequency, to see how it could improve our training program. We wanted to increase our knowledge and improve our skills, because having well-trained firefighters is vital in keeping our community safe. We learned the following.

Approximately 75 percent of our calls for service are emergency medical in nature. Similarly, 25 percent of calls are fires and false alarms. Traditionally, we focus most of our training on fire response (low-frequency, high-risk events), which is good and prudent when we consider the complex risks of our fireground operations. There was an inverse relationship in the percentage of time spent in training: 76 percent on fire topics and 24 percent spent training for EMS. The concern, ultimately, was if we were learning the lessons our incident data were trying to teach us.

As we drilled down into the data, we wanted to know more about these incidents, such as the actions taken and the outcomes. Were we, through our training program, better prepared for the incidents most frequently seen? Could we find and quantify our service delivery deficiencies and then retrain to produce a better service? Did the elements of our training program guarantee that we were growing in knowledge and increasing in skill? These may not be the same questions you are asking of your organization, but we are all asking questions of our departments that need answers.

We learned that our then-current records management program couldn’t always produce the information we needed. When it did, oftentimes the information would be outdated by the time we produced the report we wanted. To correct this issue, we found an affordable Web-based data-analysis intelligence system. This software allowed us not only to excavate but also to drill down into our information in real time. It allowed us to see our compliance with National Fire Protection Association 1710, Standard for the Organization and Deployment of Fire Suppression Operations, Emergency Medical Operations, and Special Operations to the Public by Career Fire Departments, in just a few clicks of the mouse. We could observe the time use of individual apparatus, geocode incidents with a mapping feature, and so much more. With this program, we easily excavated and interpreted our data for the first time. Such supplemental technology is vital to making decisions that improve overall service delivery.

Since we can more accurately measure our operational outputs, we could better manage their inputs. In a data-driven training program, we can apply national standards and other research contextually. Contextual implementation does not give license to dilute its value but the opportunity to improve current training criteria according to local service delivery goals. We can ensure that, throughout our department, personnel have the knowledge and skills to perform their jobs well and thus remove some of the variables that lead to injury, death, and property loss.

Our data analysis of the training program resulted in three key objectives:

  • Increase the suppression staff skills and knowledge by contextually adopting national standards and evaluating using new performance matrices.
  • Ensure a more core-competent training program according to Commission on Fire Accreditation International standards.
  • Increase the knowledge and the ability to make strategic decisions about current and future department operations through data analysis.

The benefit of being early adopters of the big data movement has not just been in analyzing isolated sections of fire department operations. Looking at the metanarrative of how fire departments do business, through data analysis, we can raise the bar of what we expect in terms of overall service delivery. Top business executives know that those who used data to drive their business decisions were, on the average, “5% more productive and 6% more profitable than their competitors.”6 Although this research concerned the private sector, we in the public sector should adapt relevant practices to improve our organizations in regard to efficacy and efficiency.

Every organization is learning to adapt to big data. The ability to measure operational inputs and outputs is a key to success. As management consultant Peter Drucker once said, “You can’t manage what you don’t measure.” We as public leaders must realize that any strategic decision made will affect our service delivery. Being a data-driven organization is by no means a “magic bullet”; rather, executive fire officers must continue to use data as the excellent tool that it is. We must let the information inform us about us.

We are stewards of one of the greatest professions in history, held in the highest regard. We must preserve this greatness.

Endnotes

1. White, Jay. (2007). Managing Information in the Public Sector. M.E. Sharpe: Armonk, New York. 1.

2. McEwen, Tom and Catherine Miller. (2004) Fire Data Analysis Handbook 2.0. FEMA: United States Fire Administration.

3. National Institute of Standards and Technology. (2011) “Fire Service Data Summit.” Technical Note 1698, http://www.nist.gov/customcf/get_pdf.cfm?pub_id=908244.

4. National Institute of Standards and Technology. (2010) “Report on Residential Fireground Field Experiments.” Technical Note 1661, https://www.iaff.org/tech/PDF/NISTReport2010.pdf.

5. Kerber, Stephen and Daniel Madrzykowski. (2013) “Fire Behavior and Fireground Considerations.” John P. Redmond Symposium. National Institute of Standards and Technology, Fire Department of New York, and International Association of Fire Fighters. http://firetrainingtoolbox.com.

6. McAfee, Andrew and Erik Brynjolfsson. (2012). “Big Data: The Management Revolution.” Harvard Business Review, October 2012, 63-64, http://hbr.org/2012/10/big-data-the-management-revolution/ar.

Looking at how departments do business, through data analysis, we can raise the bar of what we expect in terms of service delivery.

WILL LINGER has been a firefighter paramedic with the Papillion (NE) Fire Department for the past six years. A graduate of the University of Nebraska at Omaha Certified Public Manager Program, he also has a bachelor’s degree in civic leadership from that university.

More Fire Engineering Issue Articles
Fire Engineering Archives

No posts to display