The airline industry is committed to reducing its carbon footprint – with a goal of eliminating nearly all CO2 emissions by 2050, just 25 years from now. With aviation responsible for about 2.5% of annual CO2 emissions – and 4% of overall global warming – the commitment to change is a game-changer but no doubt also challenging.

To realize its commitment, the industry is going to have to rely on Sustainable Aviation Fuel (SAF), a biofuel commonly made from oil or sugar producing (ATJ) crops, including corn and sugarcane, as well as other inputs like agricultural residues, used cooking oil, and animal fats, to replace traditional fossil-based jet fuel. While the benefits of SAF are clear, in 2023, just 0.2% of all fuel used in aviation was SAF. This is mainly due to the high cost of the fuel—-the price of SAF is between three and five times the cost of traditional jet fuel— and the limited availability of SAF. This shortage, and resulting high cost, is due mainly to shortages of SAF feedstocks. As more airlines seek to meet more stringent sustainably regulations, especially in the EU, the price of SAF is on track to continue to rise. 

In order for SAF to meet its potential and become a viable, widely-used solution, significant changes and innovation are needed at the very root of its production: Harnessing AI-driven agtech to improve and lower the price of the crops that make up the fuel’s feedstock is emerging as one of the most promising ways to increase the viability of SAF as a wide-scale solution. 

Computational biology, including AI systems, is helping develop and produce seeds for feedstock crops that can be grown at scale easier in more places, even as climate change poses challenges to crop yields. Tailored seeds can also help maximize the characteristics of the crops, making them more efficient sources of SAF. 

AI systems can analyze vast datasets that include the genome, growth history, and environmental requirements for the corn, soy, and other crops currently used to produce most feedstock. Using this analysis, researchers can then develop strains of feedstock that will contain the characteristics most suited to SAF production, including low carbon impact for farming and growth  and the crops with the highest biomass produced per unit of land. Studies show that corn stover, energy sorghum, miscanthus, and switchgrass best answer these requirements. 

AI can also help analyze which areas are best  for growing these feedstocks. Studies show, for example, that crops like switchgrass, biomass sorghum, and giant reed would do well on marginal lands in the Continental climate zone. In contrast, others like Ethiopian mustard, cardoon, and castor could thrive on marginal lands in a Mediterranean climate.

In addition, researchers can use AI to engineer seeds to produce crops that require fewer resources or are tailored for certain climates and environments. Such seeds could enable crops to be grown with less water or on land that is not suitable for other crops. By analyzing the genomes of various crops to determine their characteristics, researchers will be able to determine which ones can be edited in order to enable their growth in marginal areas unlikely to produce nutritious food but could produce crops that contain specifically the characteristics that would enable efficient SAF production. For example, researchers are developing gene-editing techniques that would increase the level of extremophiles - enzymes that allow organisms to thrive in extreme environmental conditions, such as high temperatures, acidic or alkaline environments, high salinity, or high-pressure , thus producing more fuel from existing stocks. In addition to making more SAF, this approach contributes to a more sustainable agricultural economy, where land and resources are used more efficiently. 

The uses of AI to improve SAF go beyond developing more efficient crops. AI can help improve the Hydrotreated Esters Fatty Acids (HEFA) process, by which most SAF is produced. HEFA is based on hydrogenating feedstock to remove oxygen from the material, with the resulting straight paraffinic molecules cracked and isomerized to jet fuel chain length. The process is expensive and requires many resources, but it is among the most promising ways to produce SAF at scale. Using AI, researchers can develop methods to increase the efficiency of this process, 

Finally, AI could dramatically increase the amount of SAF by tapping into one of the best-unrealized sources of SAF resources – trash and food waste.  In fact, a UK study shows that the reduction in carbon emissions would be five times greater if organic garbage sources were used to produce SAF rather than incinerated. AI can help with different parts of this process; organic waste, like food waste, paper (pulp), sludge, and other compostable items, are the best candidates for SAF production, and AI-based sorting systems could more efficiently and quickly separate this valuable trash from plastic, which is less amenable to SAF conversion. 

 The changes in the airline industry could profoundly affect dozens of industries. But competition for SAF feedstock is growing – and if airlines intend to stand by their pledge (and comply with the coming regulations) while keeping fares at a reasonable price, new sources of more efficient feedstock will be needed. AI can enhance this production – and with that enhancement, all of us will be able to breathe a little more freely.