Air Force aims for military enterprise scale AI application

If artificial intelligence (AI) and robotic process automation (RPA) were wings of air, they would be flying high. For the Air Force, no less than for the other armed forces, the question is not whether to adopt these technologies. There you should adopt them first so that they get the greatest return.

“We’re in a lot of different businesses,” said Jay Bonci, Air Force chief technology officer. These include traditional air force missions, space operations, weather modeling and…

CONTINUE READING

If artificial intelligence (AI) and robotic process automation (RPA) were wings of air, they would be flying high. For the Air Force, no less than for the other armed forces, the question is not whether to adopt these technologies. There you should adopt them first so that they get the greatest return.

“We’re in a lot of different businesses,” said Jay Bonci, Air Force chief technology officer. These include traditional air force missions, space operations, weather modeling and forecasting, human performance and health planning, and financial management.

“Each of these areas is a data-rich environment,” Bonci said. “We’re working right now to make sure we can get that data into our big data platforms so it can be cleaned and indexed and ready for different communities of analysts.” He added that a lot of the data for “fusion, for things how AI and machine learning models” are intended.

Bonci spoke at a Federal News Network panel to discuss AI and RPA in a defense context.

Referring to the field of information technology, Dr. Joseph Schafer of the National Defense University, a professor in the College of Information and Cyberspace, that IT resource provisioning and provisioning is ripe for RPA. In particular, detecting and responding to cybersecurity incidents requires AI, Schafer said, due to the volume of attacks and the log data that security operations centers must deal with.

“DISA [Defense Information Systems Agency] is hit by a billion and a half cyber incidents every day,” said Schafer. “There is no way to deal with such incidents without using artificial intelligence.”

Regardless of the specific AI/RPA application, in many ways these technologies are about people. Melissa Long Dolson, IBM’s vice president of global technology sales for AI Ops and Integration, said that AI “actually brings human intelligence into the automation process. How can we help take something I learned and did all day at my job for the last 10 years and put it into a tool that I can use to do it faster, faster and better?”

In particular, when AI applies to human resources management and development, Dolson says AI systems must ensure the protection of personal data at all times.

Beyond PII, she added, “I think the DoD agencies in particular have a bigger challenge … it’s about all of these secrets that we need to protect to make sure we’re managing our national security interests the right way.”

Injecting human knowledge into an AI program, panellists emphasized, increases a person’s decision-making ability, but does not replace it. For the armed service, Bonci said, ethics center on the kill chain. Ethical use of AI maintains human judgment in fire decisions and targets. This ensures that decisions about killing chains remain traceable and comply with international conventions and US military standards.

“There are many different dimensions around this concept of AI ethics,” Bonci added. “It applies to weapons, it applies to the human factors, it applies to personnel data and how we use and combine the different sources.”

Another dimension, he said, is different levels of data classification, with the challenge being to bring together high- and low-level data for AI and analytics purposes.

“There are many efforts in the [Office of the Secretary of Defense] level by looking at how we do our data classification and privacy protection,” said Bonci.

Ethical considerations and the killing chain are even part of the NDU curriculum. “At the end of the day,” Schafer said, “no one is for Terminator,” Schafer said, referring to the Hollywood version of a robotic killer with autonomous intelligence.

Choose your AI battles

As there is no shortage of processes that could benefit from AI and RPA, it may be prudent for an agency to get an objective, third-party view of their workload in order to find the best candidates.

To help clients get where to go first, IBM consultants will examine and evaluate workloads to determine “where these AI solutions and robotic process automation solutions will benefit them most quickly, in a way that treats the data right.” , and is ethically justifiable after consideration.”

Bonci said the Air Force is essentially looking at large, high-yield applications first.

“The places where we need to think about applying AI and ML, and the places where we’re taking that path, are all areas that touch on the true scale of the military,” Bonci said. This ranges from AI for the management of weapon systems to the challenges of cyberspace.

“There will never be enough people who can respond to an automated attack,” Bonci said. “To meet this challenge, we need automated defenses that can make sense of highly noisy, highly variable traffic with the intent to confuse and deceive.”

Bonci added that the Air Force has set up an enterprise service to curate, secure and fuse data at scale for training purposes. He pointed out that the Air Force resembles commercial airlines in some functions, such as fuel optimization, efficient packaging of aircraft, and crew scheduling. Therefore, at least outside of the kill chain, industry and the military can share training data and AI approaches to common challenges.

Dolson said this type of cross-domain collaboration can help on two fronts.

“We need to share the air across industries because I think that’s a way to enable AI in your organization,” Dolson said. “But it’s also a way to get a handle on the costs that come with AI.”

For the Air Force, AI and RPA are components in ongoing digital transformation efforts. Therefore, according to Bonci, government agencies, including his own, need good policies for cloud computing, identity management, non-personal access, and secure application programming interfaces between systems. And of course the human skills needed to make everything work.

Additionally, Dolson said, the modernization “starts in one place. And that really means understanding your processes. Because if you automate a bad process, it’s still a bad process.” She said it’s better to get processes right, automate them, and then scale them on cloud infrastructure.

Comments are closed.