“The patent protects a new method of mapping and evaluating the tactical decision-making process,” said Shamim Khan, professor of computer science and head of the development team from CSU’s TSYS School of Computer Science.
The method uses visual representations of real-life scenarios, called cognitive maps, to display complex connections between decision-making factors.
“Cognitive maps have the potential to be useful decision support tools because of their ability to model a scenario consisting of concepts and events and the causal links among them,” said Khan.
The maps are used to diagram a tangled web of decision factors into a tangible portrayal of possible consequences of an action. The method presents an opportunity to improve the decision-making capabilities of military or emergency personnel from the office or classroom instead of the battlefield.
The use of the maps themselves, which are charts made up of bubbles connected by arrows, is not a new concept. Rather, Khan’s method is new because it accounts for more ambiguous or “fuzzy” cause-and-effect relationships between decision factors. For example, when a change in factor X “moderately” (or “slightly” or “strongly”) increases factor Y.
“People often want exact values, but in the real world, in real-life situations when decisions are being made, exact numbers are not always practical,” said Khan.
The “fuzziness” of the maps are what make them unique to the scientific community, and, now, to the commercial market.
The CSU team led by Khan originally developed his prototype as a military training tool with extensive collaboration from John Fuller, a retired U.S. Army colonel with 27 years of service experience.
“Most new platoon leaders will report to their first unit of assignment with little, if any, real experience as tactical leaders, yet, they command units that historically come into the most contact with the enemy,” said Fuller. “And, where chaos pervades the battlefield, where combat tends to overwhelm the senses and distort reality, leaders with the greatest amount of training tend to make the best decisions.”
Using a mix of video footage and animation, Khan’s team developed a series of simulations that progress as trainees choose from a list of available actions based on likely battlefield scenarios. After running the simulation, the actions and their consequences are mapped out and the decision made by trainees is assigned a quality score. During an after-action review, the trainee’s quality score is compared to a benchmark score representing the best possible outcome. Areas for improvement then can be identified. Courses of action leading to best possible outcomes are determined by a subject matter expert, usually an experienced military officer or war-trained veteran.
“Entire battlefield scenarios, such as convoy operation and reconnaissance patrol, can be modeled using a collection of cognitive maps,” said Khan.
Khan and a team of CSU professors, postdoctoral researchers, graduate and undergraduate students from CSU’s TSYS School of Computer Science helped develop the first prototype. The original project was funded by a $1.6 million Department of Defense grant awarded to the university in 2010.
Under the protection of a patent, Khan is working to find government or private organizations interested in developing a similar system to meet their own needs.
“Although the simulation system was designed and developed specifically for military training, a similar system would be highly applicable to other areas, including homeland security, law enforcement and emergency response,” said Khan.
RSS