Simulation science: The key to understanding national infrastructures

TRANSIMS can show traffic down to street-level details.

LOS ALAMOS, N.M., MARCH 2003 -- The continuity of American life depends on reliable operation of a complex web of interdependent infrastructures. Disruptions in any one of these could jeopardize our society. A collaboration between Los Alamos and Sandia National Laboratories is helping policymakers understand how the nation's infrastructures work, are linked, and can be protected.

Since September 11, 2001, homeland security has moved to the forefront of domestic issues. Crucial to homeland security is the continuous operation of a complex web of interdependent infrastructures such as transportation, the public health system, and telecommunications. While it is easy to appreciate that these systems constitute the basic fabric of our society, understanding their complexities is much more difficult.

Yet that understanding is being required of the nation's newly created Department of Homeland Security. The White House National Strategy for Homeland Security specifically cites the need for state-of-the-art, high-end modeling and simulation of the nation's critical infrastructures. Meeting that need requires innovative analytical tools, many of which are being developed through collaborative work between Los Alamos and Sandia National Laboratories on modeling, simulating, and analyzing the infrastructures and their interdependencies.

Chris Barrett, a simulation scientist, leads the Los Alamos research and development effort to provide the new department with tools for assessing the complex sociotechnical systems that constitute infrastructure. Developing these advanced technical capabilities is the mission of the National Infrastructure Simulation and Analysis Center (NISAC), the partnership between Los Alamos and Sandia. "Our policymakers need new simulation and information-sharing tools to help them represent and understand complex, interdependent infrastructure systems and to support decision-making for better planning, monitoring, and response to disruption," comments Barrett.

The challenge is daunting, similar to that of simulating the performance of a nuclear weapon but more difficult in certain aspects. Laws of physics determine the performance of a nuclear weapon. In contrast, sociotechnical systems are affected not only by physical laws but also by human factors--such as the decisions and actions of individuals, private industry, governments, and regulatory agencies.

Simulation Capabilities

Powerful simulation tools are the only adequate analytical means for representing and assessing this kind of complexity. Los Alamos NISAC researchers are developing infrastructure-simulation capabilities that are firmly based on contemporary mathematical and computer science theory and designed for use on modern, high-performance computing platforms. These simulations incorporate census and mobility data for entire urban populations. "We can represent systems much larger than any previously modeled by other methods because our simulations can be scaled up to include millions, or tens of millions, of interacting variables," says Barrett.

NISAC simulations are unprecedented as well because they are not confined to using aggregated (summed) values for key variables. In contrast, they are designed to use urban population-mobility data that are disaggregated to the level of individuals on a second-by-second basis. Such simulations let analysts search the data by demographics, activities, time of day, or any other parameter of interest.

NISAC researchers have already demonstrated their unique approach by developing prototype simulations for two infrastructures: transportation and public health. The first, TRANSIMS (Transportation Analysis and Simulation System), has been used in Albuquerque, Dallas, and Portland, Oregon.

Analyzing Transportation

With TRANSIMS, researchers use census data to create a statistically valid, synthetic population of millions of travelers. Drawing on such data, TRANSIMS modules assign daily activities and routes for each traveler and insert these variables into the urban traffic network. A microsimulation moves travelers to their destinations on the basis of these activities and route plans.

Urban traffic patterns emerge from the microsimulation. Automated feedback between the modules and the microsimulation accounts for travelers' responses to traffic congestion and continues until all travelers reach their destinations. Thus, TRANSIMS simulates a region's traffic movement over the transportation infrastructure and the traveling patterns of virtual residents. The result: planners can better estimate the effects of modifying the transportation infrastructure, from simply adding a new traffic light to building a new transit system.

Analyzing Epidemics

A derivative of TRANSIMS is the Epidemiological Simulation System (EpiSIMS), which couples the transmission of disease with the mobility of urban populations as developed by TRANSIMS. By coupling disease and epidemiological models to population-mobility simulation data, researchers have developed an urban simulation that helps planners understand how a disease spreads during an epidemic, what resources would be required to respond to the epidemic, and how effective various response strategies would be.

Infrastructure Interdependencies

In the real world, no infrastructure stands alone. Suppose, for example, that electrical power distribution is disrupted. When the power goes off, traffic lights go off as well, causing traffic jams. Thus, the electric power and the transportation infrastructures are linked and interdependent, compounding the complexities researchers face in analyzing these large, national networks.

Here, too, TRANSIMS has introduced new capabilities, because the traffic-light system is explicitly part of the urban traffic network module in the TRANSIMS design. Researchers can insert the disruption of traffic signals from whatever cause, and TRANSIMS will simulate the resulting behavior of travelers. EpiSIMS and other infrastructure simulations are similarly designed to represent the interdependencies among infrastructure sectors.

Today, Laboratory NISAC scientists are leveraging their expertise by extending their modeling, simulation, and analysis capabilities to other critical systems such as the energy and the banking and finance infrastructures. They have created an evolving and interoperable set of new analytical tools that are essential to the support of the Department of Homeland Security.

Says Barrett, "Our focus is on providing three capabilities to the department's functions of information analysis and infrastructure protection: studies of key technical issues, new tools for assessing infrastructure systems, and external outreach to help ensure that our work is responsive to the needs of our users."

Los Alamos National Laboratory is operated by the University of California for the National Nuclear Security Administration (NNSA) of the U.S. Department of Energy and works in partnership with NNSA's Sandia and Lawrence Livermore national laboratories to support NNSA in its mission.

For more Los Alamos news releases, visit World Wide Web site www.lanl.gov.

By Judith Machen