With AI’s Help, Knowing Every Airplane’s “Fingerprint” Can Decrease Emissions

The global aviation sector accounts for roughly 2% of energy-related CO2 emissions, according to the International Energy Agency, a figure that is rising with the rebound of post-pandemic air travel.

Decarbonization is on every airline’s radar, but the sheer complexity of the industry makes the task difficult. Flight operations comprise a complicated network of hundreds of interconnected elements – aircraft and airports, personnel and passengers – that change in real time, each triggering chain reactions.

Trying to incorporate sustainability into this sector only adds to this complexity. However, the sheer volume of data available offers the opportunity to identify areas to improve sustainability, and layering artificial intelligence over existing data allows a new level of visibility into fleets. Each individual aircraft’s mechanical and performance data are as unique as a fingerprint, and keeping on eye on variables like every planes’ maintenance schedule, passenger capacity and weight can help in assigning them to optimized routes, saving fuel.

Every A321 or B747 airplane shares the same physical traits, but no two planes perform exactly the same way. Using AI to understand those differences is a useful lever to decarbonization efforts, helping to make small efficiency gains that add up when scaled across a fleet of thousands.

Under so-called “manual” conditions, only the major factors can be considered: How many passengers can a plane carry, for example, or how far does it need to travel? But many other important factors are tail-sign dependent and quantifiable with the help of digital technology.

What is the aircraft’s engine status and age? (Due to normal lifecycle attrition, a newer aircraft is up to 30% more efficient.) How much cargo can it carry on a short-haul flight? (A huge profit driver for airlines.) What is the plane’s drag? (Believe it or not, an aircraft that’s recently had its exterior cleaned is 1% more fuel-efficient.)

With an understanding of these variables, the fuel burn can be determined. Less fuel burned reduces emissions, while performance goes up.

Time is the costliest factor in any system. Aviation datasets are continuously changing, and a human would need many hours to analyze them. But AI can distill that analysis in minutes, helping to replace traditional, sequential decision-making with a simultaneous look at system-wide operations. All this helps to ensure that tomorrow’s pilot will fly the most efficient plane available for their route.

Some in the industry are already seeing positive results. Lufthansa Group, which launched an AI program with subsidiary Swiss International Air Lines in July 2021, reports that it has saved approximately 2,000 tons of jet fuel and 8,700 tons of CO2 per year by leveraging data-driven tail optimization. That’s the equivalent of an A350 flying 16 round-trip flights from Munich to New York – all from identifying small opportunities for unseen efficiencies.

SWISS accounts for roughly 15% of Lufthansa Group’s flights. Scale results like these across the entire company, and the potential emissions and cost savings are significant. And AI is just one of many digital tools that can be used to identify operational improvements. Predictive maintenance, automated slot allocation and many other innovations help contribute to emission and cost reductions.

Together, they will play a significant part in helping Lufthansa Group meet its ambitious target to cut net carbon emissions in half by 2030 (compared to a 2019 baseline) and achieve carbon neutrality by 2050.