close
close

Apre-salomemanzo

Breaking: Beyond Headlines!

AI Challenge Powers Real-World Solutions and Workforce Skills | Article
aecifo

AI Challenge Powers Real-World Solutions and Workforce Skills | Article




AI Challenge powers real-world solutions and workforce skills



Bulletproof plate used in the AI ​​challenge which aimed to improve the fault analysis of soldiers’ bulletproof plates.
(Photo credit: US Army)

SEE THE ORIGINAL

ABERDEEN PROVING GROUND, Md. – The U.S. Army Test and Evaluation Command crowned the winner of its second annual ATEC AI Challenge at its 3rd Annual ATEC Data Summit Sept. 17.

The goal of the challenge was to engage ATEC staff to find opportunities to solve real-world problems with AI and machine learning solutions with a high potential ROI. Additionally, ATEC sought to learn how to validate testing and evaluation procedures for AI-based systems by evaluating challenge submissions to determine the winner. In turn, ATEC will apply these lessons to improve how it tests and evaluates hardware systems that integrate AI and machine learning.

This year’s AI challenge focused on how to improve the analysis of defects in soldiers’ body armor, which is currently a manual process carried out by an analyst reviewing x-ray images. By leveraging AI algorithms, ATEC saw the potential to significantly speed up this process, giving the analyst time to investigate the root cause of any identified flaws.




AI Challenge powers real-world solutions and workforce skills



X-ray image patch used during competition. The image has a defect, which can be seen as the blemish on the X-ray image of the body armor shown in Figure 1. The goal of the AI ​​was to determine if there was a defect in all x-ray images captured.
(Photo credit: US Army)

SEE THE ORIGINAL




AI Challenge powers real-world solutions and workforce skills



X-ray image patch used during competition. The image has a defect, which can be seen as the blemish on the X-ray image of the body armor shown in Figure 1. The goal of the AI ​​was to determine if there was a defect in all x-ray images captured.
(Photo credit: US Army)

SEE THE ORIGINAL




AI Challenge powers real-world solutions and workforce skills



X-ray image patch used during competition. The image has a defect, which can be seen as the blemish on the X-ray image of the body armor shown in Figure 1. The goal of the AI ​​was to determine if there was a defect in all x-ray images captured.
(Photo credit: US Army)

SEE THE ORIGINAL

Ultimately, 29 teams and 153 participants from the Department of Defense acquisition community entered the competition to develop a solution. The team’s judging criteria was based on a scoring metric that focused on detecting every flaw in the body armor. This year’s winner was the No To Destructive Testing team from the US Army’s Aberdeen Test Center.

Steven Shrewsbury, a mechanical engineer and one of five members of the winning team, explained the competition’s winning submission.

“We explored different AI models and finally chose the Vision Transformer, which is a state-of-the-art model for image analysis,” he said.

Leveraging the Data Mesh

“The AI ​​Challenge demonstrated ATEC’s ability to solve its own problems with ATEC data,” said Maj. David Niblick, Directorate of Analytics, Artificial Intelligence and Intelligence military evaluator. digital engineering within the US Army Evaluation Center, a subordinate organization to ATEC. “Teams across the acquisition community have leveraged the ATEC data mesh to provide an advanced, common computing environment to acquire and implement AI skills in a way that ATEC can put into production to save a lot of time and money. »

Niblick said the ATEC Data Mesh is a cloud-based infrastructure for housing all ATEC test and evaluation data. The data mesh integrates ATEC under one IT roof and contains modern tools and infrastructure to enable digital transformation.

However, “ultimately, ATEC relies on champions to drive the adoption of more efficient and effective methodologies,” he noted. “The 2024 AI Challenge is a phenomenal example of champion and infrastructure integration. »

This year’s competition notably saw significant growth over its 2023 counterpart, which focused on improving the analysis of acoustic sensor data for munitions testing at the Yuma Proving Ground, or YPG. The winning solution is currently being used in the YPG, saving the army hundreds of man hours.

This competition was open only to AAIDED and, therefore, only three teams participated. But by using the data mesh’s cloud infrastructure, AAIDED was able to expand the competition to a DoD-wide effort in 2024.

A collaborative environment

The ATEC Data Summit, now in its third year, proved to be the ideal venue to crown the winners. Data Summits are opportunities for engineers and scientists to come together to collaborate, share ideas and dialogue regarding progress toward digital transformation within ATEC and across the Army. They come from Brian Kelly, ATEC’s command data and analytics manager, who had the vision to digitally transform ATEC’s test and evaluation community.

Mr. Young Bang, Principal Assistant Secretary of the Army for Acquisition, Logistics and Technology, or ASA (AL&T), presented the award to this year’s winning team, alongside Mr. James Amato, Executive Technical Director of ATEC.

Mr. Bang concluded by congratulating the team and stating his intention to partner ASA (AL&T) and ATEC to expand the AI ​​Challenge to more participants and problem sets in years to come.