IMPACT OF PERFECT INFORMATION

BY RONALD DUBOIS, JAMES T. BYRNE, AND CHRIS L. SPURLOCK

Perfect” information has a significant impact on effective incident management decisions fire department incident commanders make on-scene, but such information does not exist in a dynamically changing emergency. The best one can hope for is a continuously updated flow of real-time (as things actually happen) information displayed in an easily understandable format. If such an information system existed, the incident commander (IC) would be evaluating the situation in a condition that could be described as “near-perfect.”

An emergency information system provides ICs access to the critical data and changes thereto occurring in real time at a fire emergency. This includes the fire’s exact location and temperature, the spread of smoke and carbon dioxide within the facility, trapped victims’ locations, and the current status of the facility’s involved area. Ideally, this real-time information should be updated within three seconds of a change and simultaneously become available to the on-scene IC and to all other key personnel in the chain of command.


1 Photos by David Kimmel, NetTalon Security Systems.

Look at this portrayal of a fire in progress (photo 1). Three men would have probably died when a fire, fueled by a flammable substance, quickly erupted and engulfed the room. They were cornered with no way out. The local fire department received the alarm within seconds, localized the source of the fire, and knew immediately where the victims were trapped. The IC quickly arrived on-scene and, using his laptop, developed an efficient plan to isolate and extinguish the fire and evacuate the victims. His supervisor in the chain of command had simultaneous access to the same real-time information describing the dynamically changing incident.

In such a scenario, the IC would develop a plan to isolate and extinguish the fire and evacuate the victims prior to arriving on the scene. By monitoring the progress of the incident in real time using an emergency information system, any decisions made and their results would be immediately available through the chain of command. Monitoring such real-time information allows decisions to be made immediately as the complex scene dynamically changes.

In this situation, the ICs receive “near-perfect” information that drastically improves their decision making and the odds of saving the lives of the incident victims and responding firefighters. Finally, consider the opportunities for training if all incidents could be recorded and replayed to evaluate and improve current firefighting procedures.

Below we review the processes associated with effective decision making in complex, dynamic, time-pressured emergency situations and analyze the impact of adding “near-perfect” information to the on-scene and off-scene commander’s decision-making arsenal. We also discuss the impact of such technology on future firefighter training. Such changes will fundamentally change the way incident managers do their jobs now and in the future.

CURRENT TECHNOLOGY

Although significant advances in communication and fire suppression technologies over the past decades have fundamentally changed the nature of firefighting operations, the human component, namely the local command activity, has remained a “human” endeavor. Today, on-scene commanders lead a team of trained individuals using specialized equipment whose efforts are coordinated through command, control, and communication processes to achieve specified objectives under conditions of threat, uncertainty, and limited resources. The command and control function exercised on-scene is crucial to successful emergency situation management.


Several studies have focused on those decision processes associated with more vs. less effective incident management. Effective incident management requires technical knowledge, knowledge of standard operating procedures, and the ability to process information under high-stress and limited-time conditions. Table 1 details the Behavioral Markers of Effective Incident Command as stated by McLennan, Pavlou, and Omodei in 2003.1

Let’s examine the potential impact of a “near-perfect” emergency information system available to an IC responding to a serious fire emergency. Such a decision support system would meet five out of six behavioral markers in Table 1: anticipation and planning, communication, workload management, reevaluation of the situation, and use of all available information resources.

EMERGENCY INFORMATION SYSTEM

In defining the nature of the emergency information system, it’s important to emphasize the distinction between the emergency information system and a classical “alarm” system. The classical alarm system simply alerts on detection, whereas the emergency information system continuously reports the changing site conditions. The emergency information system provides the first responders and IC with the real-time information they need to efficiently manage the scene.


2

The IC needs to have as much information as possible to gain situational awareness of the changing events at the scene. However, simply having more information is not necessarily better; it must be presented in a format that creates focus rather than confusion. Using visual tools to organize the information into a relevant format (e.g., floor plans, alarming icons) is an ideal method of reducing the complexity of the information and would relieve the IC of the task of envisioning the site conditions (photo 2).

Icons would represent the current state of individual sensors, which change dynamically as individual sensors change state during the emergency. Individual sensor icons depend on the type of sensor being displayed (e.g., security, fire, temperature, or pressure) and on its current state. The changing icons enable the incident commander to observe simultaneous alarm events, to differentiate isolated alarm events from those indicated by multiple co-located alarming sensors (smoke, temperature, carbon monoxide devices), or to discern if there are multiple intruders within a facility (motion detectors).

Not only do the icons identify alarm locations, but they also note those that were most recently in alarm (nonfire) mode. When an unacknowledged sensor returns to normal from alarm mode, an in-alarm icon is displayed indicating the recent unacknowledged alarm, as well as the sensor’s current status, giving the IC a visual “footprint” of where recent alarm activity has occurred.

Finally, the system specification should require all sensor changes be displayed on all monitoring screens within seconds of occurring.

ANTICIPATION AND PLANNING

In such a system, the IC would have the facility’s floor plan and the relative location of all sensors. Sensor status changes would be displayed in real time, providing the IC with each sensor’s status. For example, the IC could monitor the extent of smoke dispersion within the facility, the exact temperature at each temperature sensor, the exact concentration of carbon monoxide at each carbon monoxide sensor, and how these values change relative to the location of the fire within the facility.

These data provide the IC with the current status of the fire and allow remote viewers the opportunity to plan alternative methods for addressing the fire should a worst-case scenario develop unexpectedly.

COMMUNICATION AND REEVALUATION

If this information were available to the entire chain of command, communication among the IC, his subordinates, and the chain of command would significantly improve. Once the IC makes a decision, he can communicate it to his subordinates with clear, concise instructions because of the reduced stress on the decision maker resulting from that person’s having “near-perfect” information with which to make the decision.

Furthermore, the entire chain of command has access to the data on the current status of the facility. When the emergency situation warrants a change in tactics, the IC and his chain of command can discuss the change before implementing it. Multiple information sources minimize the chance that the IC will make an erroneous decision, since it is based on the best and most recent available information.

In this way, the entire emergency decision process improves because all decision makers access the same information and can evaluate how well recent decisions have impacted the status of the emergency.

GOOD INCIDENT COMMANDERS

In general, good ICs know what to look for and what to do when they find it. Rasmussen in 19832 stated that this ability does not come from years on the job but rather from the IC’s ability to reflect at length on the effectiveness of his past performance and develop a mastery of his craft through extensive rule-based decision making. Such learned rules allow the IC to use recognitional decision processes rather than slow, vulnerable analytical problem-solving processes. McLennan in 20033 noted that networks of learned rules enabling the use of recognitional decision processes form the basis of what Adams and Erickson in 20004 characterized as procedural expertise.

Unfortunately, many emergency situations are sufficiently complex to preclude using simple recognitional decision processes. Characteristics include one or more of the following: novelty-the officer never encountered such a situation before; opacity-the required information was not available; or resource inadequacy-the resources currently available weren’t sufficient to permit an optimal response.

In such situations, good ICs can transcend their limited range of specific past experiences and use fast, robust analogical decision processes to apply previous learning to novel situations. In other situations, characterized by high levels of uncertainty, ICs were forced to use analytical knowledge-based, problem- solving processes to choose an option from among a set of alternatives.

Under such circumstances, good commanders used a small number of simple and robust heuristics to guide rapid decision making about what actions to take. Adams and Erickson concluded that analogical decision processes and simple heuristics may well form the basis for adaptive expertise.

In each instance, the IC is forced to depend on his experience base and ability to develop procedural processes from those experiences to successfully manage each “novel” incident. How much better would it be if ICs received sufficient real-time information about the fire to allow them to be trained on developing procedural expertise based on the analysis of actual case studies derived from a database of actual fires where detailed archived incident information were available for review, replay, and analysis? Such a database would include a detailed time stamp for each alarm event as well as the type and value of each sensor during the incident. These data would be displayed on a graphical screen detailing the facility’s layout.

IC training could be expanded to support the development of procedural expertise based on the deliberate analysis of all available information. Over time, the archival database would become the basis for developing an “expert system” available online to all ICs and their chain of command. A simple search of this database would result in a series of “what-if” scenarios that could be used in support of novel incidents, thus satisfying the anticipation and planning criteria stated by McLennan et. al. (3) as representative for effective incident command. Incorporation of this technology in the IC training curricula would bring each trainee a procedural expertise analogous to the simulation systems now used by government and private training academies.

Furthermore, this information could be incorporated in a DVD-based training curriculum designed to provide ICs with an on-hands, cost-effective training tool for the development of procedural expertise. In effect, this training tool could be developed for use at training academies as well as local fire departments. In this way, the benefits of “near-perfect” information would be available to all fire commands for use in developing and maintaining firefighters’ incident management capabilities.

TOMORROW’S TECHNOLOGY TODAY

Over the past five years, NetTalon Security Systems, a small engineering firm in Fredericksburg, Virginia, designed and patented the NetTalon System 3000. In the past year, several System 3000 Beta sites have been operating in the Virginia region. A recent installation at the Louisiana State University Fire and Emergency Training Institute in Baton Rouge demonstrated its efficacy to the fire community.

The System 3000 reports alarm conditions to all authorized monitoring stations within two seconds of a sensor or smoke detector going into alarm. Sensor and detector conditions depicting the nature of the evolving emergency are reported immediately on a graphic representation of the building’s floor plan. Icons representing the various sensors (heat sensors, duress buttons, and smoke detectors) are overlaid on the floor plan and change color to indicate alarm conditions. The heat sensor icons display the changing temperature in real time; the smoke detectors inform the first responders of the amount of restricted visibility and potential breathing difficulty; the duress buttons display the locations of personnel trapped by the fire emergency; and the temperature sensors report the source of measured heat, leading the responders to the source of the fire.

The primary advantage of the new system is its ability to network the local fire department with the building it protects. Firefighters can examine details of a fire and begin to form their strategies before they leave the firehouse. The system’s speed and its ability to visualize the fire-involvement situation should greatly increase speed of deployment and efficacy of response.

A national study recently concluded that the average arrival time for a fire company is more than six minutes following alarm validation. System 3000 saves much of this time because firefighters arriving at the scene will already understand where the fire is, where it’s spreading, where the victims are, and with a deployment plan already formed and understood. Hence, fire departments can now envision a day when they respond to emergency situations with “near-perfect” information about the incident to which they are responding.

LSU TESTS

In a preliminary real-time test series at Louisiana State University’s Fire and Emergency Training Institute, a preliminary test series of six fires was started and monitored.

The fire department from St. George, a suburb of Baton Rouge, responded to the test fires conventionally, while LSU Fire Institute personnel and senior commanders from the St. George company monitored the fire scene remotely using System 3000 technology.

From the time each fire was started, LSU control personnel allowed two minutes to pass to simulate the time it takes to process an alarm signal. Meanwhile, System 3000 monitoring stations were viewing smoke activity at multiple points in the building within a few seconds of the fire’s starting. Within the next 30 seconds, monitoring personnel could view actual fire activity and victim dummy location. All data were received a full 90 seconds before the staged fire company received the dispatch.

After dispatch, it was another 412 minutes before the real-time firefighters located the test fire and victim dummies. Control personnel watching via System 3000 knew the fire’s intensity, location, involvement in the building, spread of smoke, victim location, and victim danger long before the fire responders even arrived. This speed of notification and remote real-time intelligence clearly demonstrated to the LSU staff and participating fire departments the advantage System 3000 can provide in fire control, victim rescue, personnel safety, and speed and efficiency of the firefighting operation.

On the operations side, System 3000 can make important fire data available to the commander and the responding crews before they leave the fire station, theoretically allowing the commander to develop his strategy and tactics, make critical decisions, and be further ahead when he arrives at the fire scene.

• • •

Fighting fires with perfect information is a dream that many fire professionals view as unattainable, since the essence of fire is chaos. Chaotic situations are difficult to quantify and define by their nature. However, the possibility of fighting fire with “near-perfect” information changes everything. Such an emergency information system is close to reality today.

We are now approaching the time where the fire industry has the opportunity to redefine the way it serves the public and emergency services by providing state-of-the-art technology designed not only to alert the first responders of an emergency but also to provide first responders with sufficient real-time information about the emergency to allow for efficient planning and response to the emergency prior to arriving on-scene.

Add to this the ability to develop an archived database of actual fire emergencies, and we enter a time when the training paradigm shifts to embrace the technology of emergency information systems.

Endnotes

1. McLennan, J., O. Pavlou, and M.M. Omodei (2003). “Cognitive control processes distinguish between better versus poorer decision making by fireground commanders.” In H. Montgomery, R. Lipshitz, and B. Brehmer (eds.). How Professionals Make Decisions. Lawrence Erlbaum Associates, 2005.

2. Rasmussen, J. “Skills, rules, and knowledge: Signal, signs, and symbols, and other distinctions in human performance.” IEEE Transactions on Systems, Man, & Cybernetics, 1983: 15, 234-243.

3. McLennan, J., M. Omodei, A. Holgate, and A.J. Wearing, “Human Information Processing Aspects of Effective Emergency Incident Management Decision Making,” presented at Human Factors Decision Making in Complex Systems Conference, Dunblane, Scotland, September 2003.

4. Adams, R.J. and A.E. Ericsson, “Introduction to the cognitive processes of expert pilots.” Journal of Human Performance in Extreme Environments, 2000: 5(1) , 44-62.t

RONALD DUBOIS is the director of finance and administration with NetTalon Security Systems in Fredericksburg, Virginia.

JAMES T. BYRNE is the chief technology officer with NetTalon Security Systems in Fredericksburg, Virginia.

CHRIS L. SPURLOCK is the coordinator for municipal fire training with the Louisiana State University Fire & Emergency Training Institute in Baton Rouge.

No posts to display