The Future of Flight: Where is Commercial Aviation Heading?

The Future of Flight: Where is Commercial Aviation Heading?

By Melita Freiberga

Introduction

As airports move further from city centres to expand and accommodate more passengers, reaching your nearest airport may take an hour by car or significantly longer if traveling on a budget via public transport. Upon arrival, you still face the airport procedures: waiting for baggage drop off or heading straight to security check, only to have your 110 ml deodorant confiscated. You then buy water and grab a snack because by now a minimum of 2 hours has already passed and you need to reward yourself. Without a deodorant, you visit the overpriced pharmacy to replace it.

Next, you might have to go through passport control. After wandering through the terminals, you eventually reach your gate ready for boarding. If everything goes smoothly, you still stand in line for about 30 minutes, wait for the luggage to be loaded, and for air traffic to clear. Once airborne, you are stuck with no internet connection. Enjoy scrolling through your photos!

After landing, you wait for the plane doors to open and navigate a maze of corridors to the exit. Passport control might surprise you with another hour-long queue. Waiting for luggage can take up to an hour, assuming it has been sent to the correct country. Finally, you take a train to get to the city, possibly followed by public transport to reach your final destination. By then, you have spent an entire day in transit, exhausted, and ready for sleep.

Current Challenges in Aviation

Sounds familiar, does it not? On top of all these overcomplicated procedures, there is a high risk of flight delays. A flight is considered delayed if it is 15 minutes later than scheduled. To give some numbers: from July 15-21, 2024, arrival punctuality in European airports was 58.4%, and departure punctuality was only 49.7%. This is an 17% increase in delays relative to the same week in 2019. Likewise, two weeks prior, arrival delays accounted for 36.3% of all European flights, compared to 29.4% in 2019.

Summer is the peak season for flying due to school vacations and favourable weather. Airports are busier, runways are congested, and staff are stretched thin. Industrial actions are more likely during this period as strikes gain maximum visibility and impact. Maintenance issues also rise with increased aircraft usage, and global warming has led to more frequent thunderstorms and floods across Europe, causing airport closures. These factors contribute to the struggle of airports and carriers to adhere to fixed timetables.

That being said, international passenger traffic across Europe is slightly below pre-pandemic levels (97% of 2019), suggesting that the aviation industry previously managed high traffic more efficiently. The primary cause for current challenges is the significant layoffs during the COVID-19 pandemic. As travel demand rebounded, airlines and airports struggled to rehire and train sufficient staff quickly. Additionally, high inflation in 2022/23 exacerbated labour disputes over stagnant wages and increased workload, leading to an outbreak of strikes in France, Italy, Germany, UK, Belgium, and other countries. This has resulted in not only increased flight delays but also cancellations.

AI's Promise: Reducing Future Flight Delays

The rise of Artificial Intelligence (AI) has led commercial aviation stakeholders to swiftly implement new tools to facilitate everyday air travel and differentiate themselves from competitors. In addition to virtual assistants powered by large language models (LLMs) and personalized marketing messages, this technology can address operational challenges. For example, machine learning (ML) services can automate tasks such as aircraft maintenance and crew planning, saving time and enhancing the customer experience.

Finnair's AI Implementation

Finnair, the Finnish flag carrier, partnered with Silo AI, Europe's largest private AI lab, to develop an ML tool that predicts potential disruptions to air traffic and minimizes the impact of adverse weather conditions on travel schedules. By leveraging historical flight data, weather information, and total runway capacity, along with current forecasts, Silo AI created a solution to assess the impact of weather on flight punctuality, allowing for advanced preparations for irregularities. Additionally, a warning mechanism was integrated to notify how many flights are expected to be delayed. This system has been further integrated with other intelligent solutions to support decision-making within the company.

Furthermore, Finnair collaborated with IBM and Apple to develop iPhone apps for their mechanics, automating and documenting processes that were previously manual. The data collected through these apps is used for predictive analytics, helping to reduce unplanned maintenance and improve overall operational efficiency.

Lufthansa's Wind Forecasting

Lufthansa Group has also been experimenting with AI. The magnitude and direction of strong wind can significantly impact flight punctuality and safety, particularly in Switzerland where the cold and dry Bise wind causes severe delays and cancellations at Zurich's Kloten Airport. Together with Google Cloud and its Vertex AI Forecast service, Lufthansa developed a wind forecasting prototype allowing the Network Operations Control team optimally schedule landings and take-offs. By analysing meteorological sensor measurements collected over the past 5 years, the team obtained insights into wind direction, speed, pressure, temperature, and humidity. After extensive data cleaning and feature engineering, the dataset was prepared for Vertex Forecast, which performs neural architecture search and hyper parameter tuning. Managed by Google Cloud, this deep learning-based approach improved Lufthansa's Bise predictions by 40% compared to internal heuristics, all within days of development.

Luggage Handling Optimization

Similarly, zeroG, a subsidiary of Lufthansa Systems, developed a solution to minimise delays caused by luggage offloading. The process began by consolidating data and identifying relevant features for predictingicting passenger connectivity – such as arrival time, arrival area, departure area, buffer transfer time and passport control flag. After thorough training, testing, and evaluation, the model was integrated within a reporting tool as well as the actual baggage permission systems. This implementation resulted in a 39% reduction in flights delayed by baggage offloading.

easyJet's Integrated Control Centre

In the spring of 2024, easyJet launched a new AI-equipped Integrated Control Centre (ICC) in Luton, featuring a generative AI tool called Jetstream. This tool provides employees with instant access to policies, procedures, and information, enabling quicker and more effective resolution of operational issues. For example, Jetstream can predict standby crew requirements and select the best crew options, thereby minimizing delays and disruption costs.

Airport Implementations

Airports are also implementing AI tools to enhance operational efficiency. AI-based software solutions designed to optimize the aircraft turnaround process and reduce delays have been adopted by Berlin Brandenburg Airport (BER) in Germany, Heathrow Airport in the UK, Schiphol and Eindhoven airports in the Netherlands, among others. BER, in collaboration with Assaia, a technology company specializing in turnaround optimization, has implemented TurnaroundControl, a system that uses computer vision event detection technology to assist airport employees in preparing an aircraft for its next flight after landing. Cameras monitor the aircraft and nearby objects, such as fuel trucks, belt loaders, and passenger stairs, and the software recommends further actions to ensure efficiency and safety. Assaia claims this system achieves an average reduction of 5 minutes in ground delays and a 17% increase in on-time performance. They have also partnered with several airlines, including British Airways, KLM, and Air France, to further enhance turnaround operations.

Conclusion

In conclusion, AI-based software is revolutionizing the aviation industry by forecasting extreme weather conditions, reducing unplanned maintenance, automating manual tasks, coordinating employees, and decreasing turnaround times. These implementations optimize operations and enhance customer satisfaction by adopting more reliable and robust methods. Punctuality remains a key focus for the aviation industry, and these innovations offer a promising outlook for the future of air travel. As technology continues to advance, we can expect even more breakthroughs that will further transform the industry, hopefully resulting in as few delays as possible.

Bibliography

Finnair partners with Silo AI to predict flight delays. (2023, March 15). Silo AI. https://www.silo.ai/news/finnair-partners-with-silo-ai-to-predict-flight-delays

Google Cloud. (2023, June 8). Lufthansa Group: Predicting wind conditions with AI. Google Cloud Blog. https://cloud.google.com/blog/products/ai-machine-learning/lufthansa-group-predicting-wind-conditions-with-ai

Assaia. (2023). TurnaroundControl: AI-powered aircraft turnaround management. https://assaia.com/products/turnaroundcontrol

easyJet. (2024, April 3). easyJet launches new AI-powered Integrated Control Centre. easyJet Media Centre. https://mediacentre.easyjet.com/story/15478/easyjet-launches-new-ai-powered-integrated-control-centre