AI chatbots, baggage screening, flight price predictions, and AI pilots are all part of the modernisation of the aviation industry.
- Chatbots improve customer service
- AI increases airport security
- Simplifying maintenance and price management
- AI pilots land airplanes
- The pitfalls of AI in planes
- Investing in rigorous safety measures
There was a time when the term artificial intelligence (AI) invoked images of humanoid robots and dystopian futures. But today, AI algorithms are an indispensable part of numerous devices and applications we use on a daily basis. Netflix, for example, uses AI to make tailored recommendations to every user and improve streaming quality. Similarly, Instagram uses machine learning to find new accounts users might enjoy and display them on their personalised explore page.
Outside the world of social media and streaming services, large industries, such as aviation, which is expected to generate global revenue of $872 billion in 2020, are also increasingly relying on AI to improve their services. Whether that involves using AI to improve the customer service experience or to predict when an airplane will need maintenance, the global AI in aviation market is expected to reach $2.2 billion by 2025.
Chatbots improve customer service
In 2020, airlines are expected to transport a record of 4.72 billion passengers. With an ever-increasing number of fliers, airlines had to find an efficient and cost-effective way of handling vast quantities of customer queries. In fact, 68 per cent of airlines and 42 per cent of airports have implemented AI-powered chatbots to provide flight information, reservation details, and answer common questions related to arrivals and departures. In addition, such chatbots are available 24 hours a day, so airlines don’t need to employ a customer service representative in their place.
Airlines are also increasingly relying on popular messaging apps to communicate with their customers. Icelandair, for example, is using Facebook Messenger as a platform for its new chatbot. Created by the airline consultancy Travelaer, Icelandair’s chatbot can answer questions, book flights, and even offer interesting facts about Iceland. And if the chatbot can’t adequately respond to the customer’s request, it quickly forwards them to a human agent. Some airlines have also employed voice assistants to provide answers to frequently asked questions. United Airlines, for example, uses Amazon’s Alexa to communicate with customers regarding their flight status and check-in requests.
AI increases airport security
Security is of paramount importance to every airport. In 2019, the AI security and defence company Synapse Technology announced the release of the first patented AI platform for X-ray machines, called Syntech ONE 200. The Osaka Airport in Japan has already ordered the platform, which makes it the first airport to introduce live AI technology for baggage screening. Syntech ONE 200 automatically detects threats such as weapons and increases the speed with which passengers move through security checkpoints.
In the US, the Oakland International Airport also modernised its security with an AI system provided by Evolv Technology, called Evolv Edge. It’s a screening platform that includes all the necessary software, hardware, and accessories, and comes with wheels that allow airport security to easily transport it from one area to another. Furthermore, its AI software platform can detect a variety of weapons at a much higher speed than legacy metal detectors.
Simplifying maintenance and price management
Good airport security, however, means little if the aircraft itself is faulty. In fact, some of the most tragic airplane disasters in history were caused by poor maintenance. In less extreme cases, maintenance is also one of the leading causes of flight delays. As a solution, companies have begun implementing AI in an attempt to predict maintenance failures before they happen. The world’s largest airplane manufacturer, Airbus, is equipping its aircraft with systems that can record large amounts of data in real time. The data is then uploaded to Airbus’ cloud-based data repository, called Skywise, which runs an analysis and makes future maintenance predictions.
Apart from maintenance, airlines must also stay on top of flight pricing, which tends to fluctuate based on a multitude of factors, such as seasonality, competition, oil prices, flight distance, and time of purchase. Considering the variability of these factors, companies must regularly update their prices. AI can simplify this process by analysing previous data and making appropriate predictions. Some companies also want to use AI to keep track of their competitors. Faculty, a British company specialised in AI solutions, developed an AI model for an airline that wanted to receive predictions on its competition’s flight prices, so it could modify its own prices accordingly. Faculty’s model was able to provide forecasts that were between 70 per cent and 80 per cent accurate up to 90 days before every flight.
AI pilots land airplanes
In the future, planes might be entirely operated by AI systems, which could even be used to perform landings. A new AI system developed by researchers at the technical universities of Munich and Braunschweig analyses visual data of the runway and adjusts the flight controls without any human assistance. It even managed to successfully land a small plane with passengers on board at the Diamond Aircraft airfield in Austria in May 2019. Furthermore, the system can detect infrared light in addition to the normal visible spectrum, which means it can function in severe weather conditions, such as fog, that can make it challenging for a human pilot to see the landing strip. It also doesn’t rely on Instrument Landing System radio signals, which is often too expensive for smaller airports. While there are concerns that false markings on the ground could trick the AI, Dr David Leslie at the Alan Turing Institute believes the risks are minimal. “Automation doesn’t necessarily mean that the pilot would be any less in control of the flight, it could mean that they’re more well [sic] supported by the technologies to land the plane,” Dr Leslie explained.
The pitfalls of AI in planes
Unfortunately, AI systems aren’t always an improvement and have even led to tragedy. One of the causes behind the two fatal Boeing 737 Max crashes in Indonesia and Ethiopia was AI software called the MCAS (Maneuvering Characteristics Augmentation System). Initially, MCAS was designed as software with limited capabilities that would only switch on to course correct the plane’s nose if the sensors detected extreme force and wind resistance. But when Boeing updated older plane models with bigger engines, it also increased the amount of control MCAS had over the plane. However, pilots weren’t aware of the level of power the AI software had or what to do if it misfired, which resulted in tragedy. On the Indonesian flight, data showed that the AI system repeatedly pushed the nose of the plane down while the pilot tried to pull it up, until the plane crashed, killing everyone on board.
Investing in rigorous safety measures
The introduction of AI to the aviation industry has simplified numerous tasks for airlines and airports. Passengers can be screened more quickly and efficiently, while airport security can use AI as an extra pair of eyes when screening baggage. Moreover, chatbots are allowing customers to book their flights and get answers to common questions easily. But malfunctioning AI software can also lead to fatal crashes, as seen in the case of the Boeing 737 Max. To avoid future tragedies, the aviation industry should be rigorous in its safety measures and training, so that pilots are well-equipped to handle any potential misfiring of an AI system.