The Air Force’s chief scientist is pursuing rapid prototyping as a means to combat costly “requirements churn.”

Mica Endsley said Wednesday the Defense Department, over the years, hasn’t done a good job defining requirements, which has cost the services money and delayed programs. Endsley said the idea of early and rapid prototyping allows DoD to make changes and work out kinks before physically building a weapons system. DF-ST-87-06962

“That’s the much more cost-effective way to do it and it also creates much more effective systems,” Endsley said during a presentation at a Women in Aerospace event in Arlington, Va.

DoD defines a requirement as the need or demand for personnel, equipment, facilities or other resources or services by specific quantities for specific periods of time, or at a specific time.

An advantage of rapid prototyping, Endsley said, is designing how a system will look, how it will operate, what its displays will look like and what its functionality will be, all without writing a single line of software code. This allows users to “push all the buttons,” Endsley said, and allow the Air Force to find out what works and what doesn’t before it is too late.

Following her presentation, Endsley said the Air Force is pushing a program within its research laboratories that does rapid prototyping with intelligence systems. PCPAD-X, formally known as Planning & Direction, Collection, Processing & Exploitation, Analysis & Production and Dissemination Experimentation, is an Air Force Research Laboratory (AFRL) initiative that streamlines the path by which multi-intelligence (multi-INT) technology transitions to the user community.

Endsley said PCPAD-X allows the service to test ideas and concepts and find out what’s really working before buying a particular technology or creating a new technology. PCPAD-X analyzes specific user groups, documenting their respective missions and the associated requirements, according to the Air Force. Each analysis yields a tailored breakdown of these missions and requirements into core problems, which are further divided into functional needs and user workflow needs.

“We have to do a better job going of going in and doing good requirements analysis and good systems engineering up front so we know exactly what we’re building before we build it,” Endsley said.