Test Professionals pitch for AI solutions at DAF-MIT ideation workshop

  • Published
  • By Staff Sgt. Kristen Pittman
  • 403rd Wing Public Affairs
CAMBRIDGE, Mass. --  

The  Department  of  the  Air  Force-Massachusetts  Institute  of Technology  Artificial  Intelligence  Accelerator  hosted  a  four-month long AI Ideation workshop spanning from November 2021 to February 2022.

The  workshop  stemmed  from  a  need  members  of  the  AIA recognized  as  a  lack  of  foundation  for  recognition  of  problems that  could  be  solved  using  machine  learning  and  the  ability  to implement AI in areas that have problems that could benefit from a better understanding of AI and machine learning capabilities.

“The intent of this ideation program is to empower Airmen and Guardians and help them along in identifying problems they have that machine learning can help solve,” said Col. Tucker Hamilton DAF-MIT Director. “We want to give them the tools and references and networking they need to help them actually solve those problems using machine learning.”

Nineteen   Airmen   and   Guardians   from   the   flight   test   and evaluation    career-field    participated    in    the    mostly    virtual workshop   spending   around   three   hours   a   week   receiving instruction from various entities including MIT’s Lincoln Lab and the  Air  Force  Academy’s  CyberWorx  as  well  as  completing lessons on Digital U revolving around human-centered design and development and AI education.

“For  this  beta  iteration,  we  went  with  the  flight  test  and engineering  career-field  because  we  recognized  they  work  with and have a lot of knowledge regarding data,” said U.S. Air Force Col. Tucker Hamilton, DAF-MIT director. “We knew they would be most receptive to this program and would be good candidates for feedback.”

Hamilton   said they   plan   to   involve   other career-fields   like operations analysts in the future.

“For this ideation cohort, we focused on three pillars,” said U.S. Air   Force   Capt.   Victor Lopez, chief of   Air   Force   artificial intelligence for persistent attack and reconnaissance with the MIT AIA Program. “Problem decomposition, artificial intelligence—the basics, and contracting.”

Space Force 1st Lt. Sean Haliyur, division engineering manager at the Air Force’s SEEK EAGLE office with the 96th Test Wing at Eglin   Air   Force   Base,   Fla.,   said   prior   to   the   course,   his background  in  AI  was  minimal,  and  that  the  coursework  was  a beneficial  in-depth  look  at  all  of  the  factors  that  come  with problem decomposition and user experience.

“I was able to work alongside a cohort from Lincoln Lab who is working to predict when there will be (foreign object debris) on flight lines  and  it  allowed  me  to  see  an  in-depth  look  at  how people  are  able to  work  through the  process  of  problem  solving to acquire data needed,” said Haliyur. “I wouldn’t have known where to start. To me it seemed like an extremely hard problem to tackle considering the data you would need and the lack of it, so  it  was  beneficial  to  see  and  walk  through  that  process  with them.”

In  early  January,  12  problem  challenges  were  presented  ranging from  using  natural  language  processing  solutions  for  unmanned aircraft  teaming  to  airspace  scheduling  to  space  surveillance,  and five projects were voted on to move forward in the development process. “We put people in teams with these five ideas to enable them to work  together  and  dig  deeper  to  understand  more  in  depth  the data and software required to bring these solutions to fruition,” said Lopez.

The  course  culminated  in  an  on-site  capstone  at  MIT  where Hamilton  said  students  were  able  to  sit  down  with  contracting officers,  acquisition  professionals,  and  machine  learning  subject matter experts to polish their ideas and develop a vetted business plan to present to their respective leadership.

As  far  as  impact,  Lopez  said  that  at  a  minimum,  he  hopes  that through  the  help  of  the  resources  from  the  workshop  and  the capstone, the participants will have a solid roadmap with which to tackle  problems  with  machine  learning  solutions  and  be  able  to share that knowledge with others. But he hopes that some of the ideas presented can actually be implemented and make an impact on operations whether base-or branch-wide.

 Long-term, he said he can see this basic ideation education being implemented  in  a  more  widespread  way  whether  it’s  through professional   military   education   courses   or   certain   technical school courses.

“I think this type of education and support is extremely important for the   entire Department   of   the   Air   Force to   empower individuals,” said Hamilton. “For our 21st Century battle space--dealing  with  emerging  technologies  and  the  digitally  empowered individual  and  force,  we  need  to  give  Airmen  and  Guardians  the tools they need to be successful in identifying how they can solve their own problems using technology.”