Overview

Where’s my package? We’ve all been there. It’s Thursday, and something we ordered for delivery on Wednesday still hasn’t shown up. Frustrating? Of course. Life-or-death? Not likely. But when we’re talking about military cargo, it often is.

The U.S. Department of Defense (DoD) transports millions of tons of cargo over air, land and sea every day. To give some context to the sheer volume: An Air Force Boeing C-17 carries a maximum of 77 tons. Just 1 million tons would require about 14,000 of these planes!

Because some of this cargo is vital to military operations worldwide, it must be delivered as fast as possible. Getting critical shipments—such as military vehicle repair parts—where they need to go can mean the difference between winning and losing a battle. Likewise, specialized medical supplies and blood products must arrive where they’re needed, when they’re needed, as lives literally depend on them.

The challenge

Millions of tons of cargo

When it comes to priority cargo, air is the fastest way to get items from point A to B. A group of Army specialists known as the Army Airlift Clearance Authority (AACA), helps determine whether and when shipping via air makes the most sense. AACA’s mission is to make sure the highest priority items, such as engines and transmissions, get sent via air efficiently and effectively. 

If the cargo meets key business rules and criteria, AACA clears the cargo for air transportation. If not, AACA staff will challenge the air cargo request and contact the shipper or customer to determine urgency of need and required delivery date. If AACA determines that the cargo request does not meet the business rules, it will divert the cargo to less costly modes of transportation, such as truck or cargo vessel.

In 2016, AACA received approximately 260,000 requests for air travel, requiring substantial time and labor hours to evaluate and meet these high-priority demands. In addition, AACA didn’t receive about 55,000 additional requests that therefore did not get processed.  

 

To help speed the AACA’s ability to process requests, the Army invested in a three-week pilot using IBM’s Watson AI services. AACA wanted to determine whether it could find less costly transportation--such as ground or sea--instead of costlier air transportation.                                                                                                                                                                                                                                                                                                                                                                                                                                    The data was comprised mostly of structured data, such as Excel spreadsheets and tables, which meant very little time was spent preparing the data. The real challenge was developing an understanding of the AACA business process and replicating that process in IBM’s AI model. Once the model was constructed, testing data samples were applied to validate the model’s sequence and output.

Transformation story

Applying AI

IBM “fed” Watson 55,000 transportation requests from 2016 (just a small portion of total requests from that year), and, in three weeks, taught the system how to interpret this data using AACA’s metrics.

“It would’ve taken AACA three to four months to process the 55,000 requests,” says Joe Corleto, IBM Army Readiness Program Manager. “It took IBM’s Watson a little more than an hour.”

 

Watson identified $26 million in cost avoidance in 2016, which would have saved an impressive 30 percent of total costs had they used AI that year instead of just humans alone. And, while the AACA found $94 million in cost avoidance for fiscal year 2016 through its normal operations, having a Watson AI capability could have generated more than $120 million in cost avoidance to the Army's second-destination transportation fund.

Results story

Good decisions save lives

One of the most important aspects of the program is its emphasis on the “augmented intelligence” that Watson provides to AACA staff. The strengths of human-machine collaboration can be more effective than either force acting alone. As such, AACA technicians reviewed and validated Watson’s recommendations.  “This is why the Army wants to automate this process by augmenting smart humans with smart technology,” Corleto says.

Following the success of the pilot, the Army approved an expansion to the project and is tapping IBM to launch the program across AACA operations. When scaled to full production, benefits include increased cost avoidance that could total up to $1 billion over 10 years.

But it’s about more than money, it’s about readiness and ensuring soldiers have what they need to win. “The pilot has helped demonstrate the Army’s ability to make good decisions from tons of data,” Corleto says. “Good decisions in the military means lives saved and missions accomplished.”