Traditional models examine a representative sample of the population, using average behavior to extrapolate findings for the whole. But while averages provide some insight, chances are slight that your organization will actually experience an “average” occurrence. And such analyses are often static, assuming entities respond in prescribed manners even as their environment evolves. Decisions are better made against a range of probable experiences.

With Agent-Based Simulation (ABS), characteristics of specific entities, or agents, are modeled. These may represent organisms, people, goods, equipment or organizations. Nothing is described in terms of averages. Rather, each agent is its own program, comprised of both data and behavioral rules (processes) that act on the data, capable of learning and decision making. Together, their cumulative action provides greater insight into emergent behavior and potential scenario impact.

ABS is particularly useful in dynamic environments, with heterogeneous populations, or if entities interact, adapt (learn experientially), or act in non-formulaic manners. Consider vehicles in traffic, insect colonies, financial markets, immune responses, consumer purchasing patterns, competition, ecosystems, intelligent buildings, and constrained supply chains. ABS allows the rapid exploration of a practically infinite range of social and environmental phenomena.

Actions taken to address one type of risk can often increase exposure to other risks. Companies need to implement integrated risk management to identify and manage interdependencies among all the risks facing the firm.