Delta Air Lines aircraft at the gate Delta Air Lines

Delta’s Predictive Analytics Drives On-Time Performance Improvements

Delta, an early adopter of data analytics, is using it to differentiate its operational performance.

SAN FRANCISCO--Delta Air Lines has avoided 1,000 potential engine events over the last 12 months largely because of its approach to asset performance monitoring and operational goals. That includes striving for 100% completion factors every day—a goal Delta achieved for 222 days last year, says Andrew Mahon, manager of propulsion engineering for the airline. The completion factor days has been rising steadily since 2010 when it was zero. Correlated with its on-time performance improvements, Delta Air Lines’ maintenance-related cancellations have dropped by 98% from 2010 to 2016 (5,212 in 2010 to 123 in 2016), says Mahon.

Delta has proactively used data monitoring tools for more than a decade—starting with SmartSignal in 2003--and added Bit Stew and GE’s Asset Performance Management (APM) system last year to prioritize alerts and allow engine performance modeling and trending, says Mahon, speaking at GE’s Minds & Machines event.

Delta is using predictive models to get the right people, tools, parts and instructions to the right aircraft at the right time to “know with confidence” that the fix will be right the first time. “We feel predictive analytics will get us there,” he says.

Delta monitors more than 1,800 engines—from its own fleet as well as those for third-party customers—and these can generate 4,000 fault codes daily. The new tools have helped aggregate the fault codes and “pareto” them to help technical analysts prioritize engines and faults, then investigate the problems,says Mahon.

Getting the codes and corrections into a system makes the fixes more institutionalized, as well.

Mahon says Delta is putting all of its engines into APM, and not just the GE powerplants, because it wanted to have an OEM-agnostic, holistic solution.

Delta operates 830 aircraft—19 types in all—that generate a range a data quantities—from MD88s that transmit 0.5 kilobytes of data per flight leg to the A350XB, which transmits 176 megabytes of data per flight leg.

TAGS: Big Data
Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.