US Army soldiers conducting combat operations in Afghanistan are faced with numerous logistics challenges. One major challenge is maintaining the complex electronic weapon systems and equipment that are used to conduct full-spectrum operations against the enemy, to protect the fighting force, and to provide critical life-support functions in remote areas. Soldiers need real-time access to the collective wisdom, expertise, and aggregated knowledge base on a given electronic system. The US Army CECOM Training Support Division (TSD) is harnessing the power of Bayesian Belief Networks (BBNs) to address this specific need and to mitigate some of the logistics challenges that are inherent to the Afghanistan theater.
CECOM Equipment Diagnostic Analysis Tool, Virtual Logistics Assistance Representative (CEDAT VLAR) is a BBN-based tool for Command, Control, Communications, Computers - Intelligence, Surveillance, Reconnaissance (C4ISR) weapon systems that encodes system knowledge and guides soldiers through diagnostic processes that are validated by experts. CEDAT VLAR is currently in use in Afghanistan supporting operators and maintainers of RF over fiber radio systems and is currently being developed for tactical power generation systems, RADAR systems, and satellite communications systems. Very little data is available for these systems so the elicitation process is key to the efficiency and accuracy of CEDAT VLAR and, also, the major development challenge. CECOM equipment experts are located all over the world. CECOM TSD is using Bayesia Lab and the BayesiaLab Expert Knowledge Elicitation Environment (BEKEE) to provide the necessary tools and functionality for our knowledge engineers to reach those experts remotely without any loss in quality of information from one-to-one interviews. BEKEE processes and analytics coupled with a custom BEKEE translator give TSD developers a unique and highly effective method for working with experts and delivering quality software to Warfighters!