Air Force turns to AI to help maintain bombers and ICBMs


The United States Air Force Global Strike Command will deploy artificial intelligence technology in a bid to increase the reliability of its nuclear bombers and ICBMs as part of an expansion of its partnership with Virtualitics, a provider of artificial intelligence and software and data exploration services.

From its headquarters at Barksdale Air Force Base in Louisiana, Global Strike Command oversees the nation’s fleet of nuclear-capable strategic bombers, including the B-1B Lancer, B-2 Spirit, and B-52 Stratofortress bombers, as well as three missile wings of the Minuteman III intercontinental ballistic missile (ICBM).

Global Strike Command was created in 2008 as the USAF’s primary command, and follow-up to Strategic Air Command, following a pair of incidents in which nuclear warheads and vehicle assemblies were mistakenly loaded onto aircraft. The entire Air Force nuclear arsenal (representing two-thirds of the US military’s total nuclear arsenal) is now managed through Global Strike Command.

Maintaining mission readiness and the ability to project power is at the heart of all US military branches, and Global Strike Command is no different. But with a fleet of sophisticated vintage aircraft and the need to manage nuclear material, not to mention the strategic nature of the mission, ensuring mission readiness takes on a unique flavor for Global Strike Command.

The Global Strike Command “Mission Capability” rate is the primary metric indicating the health and readiness of an aircraft fleet. The USAF, along with other military branches, has been trying to improve this metric for years, but has in fact been losing ground.

A 2020 report from the Government Accountability Office highlighted the main problems of preparation with aircraft throughout the army. None of the three bombers that make up the air component of the nuclear triad earned a passing grade. The B-2 Spirit, a relative baby with just 25 years of fleet service, met mission capability rate in six of 11 years from 2011 to 2021. The B-52, which the USAF began flying in 1954 , met the mark. for three of the 11 years, while the B-1B, which entered service in 1986, only did so once.

From bottom to top, B-2s, B-1Bs and B-52s fly in formation over the Sierra Nevada (image courtesy of the USAF)

Almost all of the planes were labeled by the GAO report as failing the mission. “We looked at 49 types of military aircraft and found that only four types met their annual mission readiness goals from fiscal year 2011 through fiscal year 2021, an overall decline over time,” the GAO wrote. The GAO report highlighted problems with the B-1B, including unscheduled maintenance and shortages and delays in procuring spare parts that worsen the maintenance backlog, according to an article from 2020 in the Air Force Times.

virtualitica began its relationship with Global Strike Command three years ago under a Small Business Innovation Research (SBIR) contract, and the relationship has evolved ever since. This week, the Pasadena, California-based company announced it is moving forward with an expansion of the partnership that will see the construction of a center of excellence with Global Strike Command as part of a plan to expand the use of AI.

The goal is to harness the predictive capabilities of AI to increase the aircraft availability and overall mission readiness of the ICBM fleet. To that end, Virtualitics will assist Global Strike Command in specific areas, including predictive maintenance, inventory management, supply chain optimization and manpower resource allocation.

virtuality counts data name which will deliver dashboards and analysis tools that USAF airmen will use directly. “We provide training for analysts who want to learn more about developing and analyzing with the platform, but no training is required to use the system,” the company says.

The system will recommend which parts should be replaced and provide “clear written explanations as to why the algorithm recommends replacing those parts,” the company says. “The models will also take into account the supply chain implications of these recommendations and maintainers’ scheduling constraints, thereby optimizing these recommendations.”

Predictive maintenance will be a game changer for commanders on the ground who keep these aircraft and ICBMs ready to get the job done, said Maj. Gen. Jeff Taliaferro, US Air Force (retired).

“Virtualitics makes it possible not only to improve day-to-day decisions, but also, more importantly, implementation decisions,” Taliaferro said in a press release. “Knowing in advance that an aircraft will need a major repair before deployment will allow for much better decisions that could save missions and millions of taxpayer dollars.”

Virtualitics develops an AI platform called Intelligent Exploration that uses AI and machine learning to make sense of data, including detecting correlations and anomalies. The software has a visualization component and also uses “plain language” explanations to help users understand what it is doing. The company has clients in the areas of life sciences, technology, financial services and government.

Related Posts:

Why AI is a critical capability for the US Space Force.

Booz Allen gives government a deep learning advantage

Tamr helps Air Force dispute data

Leave a Reply

Your email address will not be published. Required fields are marked *