Artificial Intelligence (AI) is steadily being adopted in a multitude of commercial products and services. It is pushing the boundaries of automation to places we could not realize before. And the Aviation industry is no different. In this article, you can read about how AI is being applied in six different domains of Aviation. From operation optimization to fully autonomous planes along with leading companies and startups.

Operation Optimization

Different workflows in the aviation industry are characterized by dynamicity and are subject to disruptions. Such disruptions cost airlines loads of money. For example, unplanned maintenance leads to delays which in turn add expenses as a result of expedited transportation of parts, overtime compensation for crews, and compensation for travels. Most notable on this front is Airbus’s Skywise open data platform which aims to improve aviation operational performance and business results. The platform has been used by easyJet for predictive maintenance. The process reduces delay-induced costs by predicting failures ahead of time.

AI not only can predict failures but also shorten the time required to find a fix. For instance, SynapseMX is an AI startup that supports maintenance teams to make quicker technical decisions. And Donecle, a Toulouse-based company, is developing autonomous aircraft inspection UAVs. They utilize the latest AI image analysis algorithms to identify defects on the aircraft’s skin.

User Experience

Airlines are using a wide range of AI technologies to provide the best experience for their customers. Delta Airlines for example is investing in technologies that enhance airport experience by providing self-service solutions. Namely, it introduced face recognition technology to confirm a traveler’s identity by matching their faces with passport photos.

AI-based natural language processing (NLP) technologies are particularly useful in user experience. Consider a case of a major disruption that leads to canceled flights and delays. The aviation company will be concerned with solving the problem, but it also has to deal with troubled passengers. Here is where AI came to the rescue. Dave O’Flanagan, the CEO of Boxever, argues that airlines can adopt chatbots to deal with passengers in case of a disruption. This will reduce the load on call centers and customer support agents.

NLP can also be used to harvest and analyze customers’ feedback. Air traveling can be stressful for many. One has to do various tasks like passing through security control, checking luggage, finding gates, waiting, and whatnot. If airlines can aggregate customers’ feedback and extract patterns of discomfort or uneasiness, they will be able to enhance customer service promptly. For instance, Southwest Airlines have established the Listening Center where a group of experts monitors social media feeds that are related to the airline, its competitors, and aviation in general. AI is used to summarize, categorize, and extract sentiment from the collected data. This operation allows the company to respond and resolve emerging issues before they develop.

Saftey

Safety measures are among the top priorities in the aviation industry. No doubt that big improvements have been done so far, and with AI more is coming. Oreyeon, a Lebanese AI startup with offices in Portugal and Switzerland, specializes in developing airport solutions with a focus on aviation safety. Its’ runway surface monitoring system enables real-time detection of FODs in the airfield. FODs, or Foreign Object Debris, include any object that shouldn’t naturally exist on the runway and may cause damage. According to the startup, FODs cost the aviation industry more than 15 Billion dollars per year regularly damaging landing gears or engine fan blades.

Aviation safety should be available for anyone. That is what Pilota, a New York air transportation startup, is trying to achieve. Pilota leverages machine learning algorithms to predict flight disruptions and automatically rebook the user ahead of time. It also provides information about safety measures that airlines are taking or are not taking.

Revenue Management

Airlines are using advanced data analytics techniques and algorithms to maximize their profits and reduce their expenses. They crunch customers’ data to estimate their willingness to pay and set dynamic pricing strategies for flights, fare class, and ancillary. They define destinations – or where to fly – by predicting demand from users’ search data. They reduce the quantity of dumped food after each flight by analyzing historical food ordering data. They also reduce CO2 emissions into the atmosphere.

The path that a plane takes depends on different factors such as weather and flight traffic. By aggregating data from different sources, AI can be used to forecast such factors and optimize the flight path. This optimization leads to a shorter flight time, less fuel burn, and eventually less CO2 emissions. However, the gain is not only ecological but also financial. A statistical report by IATA in 2012 concluded that airlines spend 33% of operational costs on fuel.

Aircraft Manufacturing

AI is even being used in designing lighter, stronger, and more efficient aircraft. Autodesk Research is working on the new Generative Design technology to make it mainstream. The technology is based on handing an AI algorithm a set of design constraints (lightweight, strong, low-cost, …) and ask it to go through an exhaustive set of choices to find the one that best fits the requirements. The designs end up looking weird and resemble shapes and structures found in the natural world.

Autonomous Aviation

While it is too early for AI to replace pilots in your typical commercial aviation. Autonomous aviation is currently limited to UAV flights and short flights (or flying taxis). In December 2019 Airbus announced the completion of an automatic takeoff test that is based only on computer vision. This was part of its Autonomous Taxi, Take-Off & Landing (ATTOL) project. Autonomous aerial vehicles (AAVs) have already entered service in China for aerial cinematography, photography, and emergency response. The vehicles are developed by EHang, a Chinese company that assembles passenger AAVs. The vehicles are powered by AI and require no pilot at all.

Adoption Challenges

While Artificial Intelligence is being developed for a wide range of applications, its adoption still faces several barriers especially in safety-critical applications such as aviation. One of the main barriers is how to build public trust in AI-based systems. To answer similarly themed questions, the European Union Aviation Safety Agency (EASA) published in February 2020 an AI Roadmap based on AI trustworthiness to tackle the ethical and societal issues around AI. The air transport system is facing new challenges: increase in air traffic volumes, more stringent environmental standards, growing complexity of systems, a greater focus on competitiveness, for which AI could provide opportunities