Summary
According to Merriam-Webster, artificial intelligence (AI) is “a branch of computer science dealing with the simulation of intelligent behavior in computers” and “the capability of a machine to imitate intelligent human behavior.”[1] AI also encompasses machine learning, computer vision, and many other technologies. (For more detail on these technologies, check out Artificial Intelligence, a Technology Focus article.)
But what can AI do for airport operators? What are some of the challenges it presents? Airport operators saw the prominence and capabilities of AI skyrocket within the last year, and they want answers to these questions. The ability of AI to learn from data and make autonomous decisions opens doors for enhancing efficiency, security, and passenger experience. Yet AI is not without pitfalls—it can make mistakes and harbor biases that could impact performance and outcomes. For instance, misidentification of passengers or baggage by AI systems can lead to flight delays or cancellations.
This article is the first of a two-part series on AI solutions, and it walks through many opportunities for using AI at airports, including both exploratory uses and available solutions. Part II summarizes the challenges that airport operators should prepare for as they incorporate AI into their operations.
Passenger Experience Improvements
Wayfinding
Although real-world deployments are a few years out, AI could be combined with Internet of things sensors and integrated into dynamic signage solutions to provide passengers with personalized information as they pass by. For example, a sign could display flight info to a passenger as they pass by or direct a passenger with a later flight to a nearby concession area. AI algorithms could also enable the airport to automatically display ads in certain areas based on the flight demographics in that part of the airport.
For example, in 2022, Delta tested a parallel reality sign at Detroit Metropolitan Airport. If passengers opted in, a kiosk and a large dynamic sign would display their personal flight information when they stood close by, while showing nearby passengers different, personalized information. The sign would detect each passenger and make their information visible from only their location, for up to 100 passengers concurrently.[2] Products such as parallel reality signs will only get more advanced as AI technology develops.
Virtual Assistant to Passengers
Similar to AI that has been used in chatbots for many years, airports could deploy virtual assistants with advanced response capabilities, whether on the airport website/app or in kiosks around the facility, easing the number of calls into the airport call center or questions asked at information desks. Over time, these features will become more advanced and able to handle more unscripted interactions with passengers.
A more advanced solution might be an AI-generated avatar on screens at information desks that “listens” to passenger questions and responds in a human-like voice. Some airports have already tested the use of screens and webcams at information desks to connect passengers with help desk staff in another location, as the Louisville Muhammad Ali International Airport did during the COVID-19 pandemic.[3]
Another type of passenger assistance, AI-powered language translation, could improve the tools available to airport staff for communicating with passengers who speak different languages. Solutions already exist to provide live translation through smartphones; in the future, tools such as augmented reality headsets could also be leveraged in these situations.
Trip Planning
Airport operators could also deploy AI-enabled solutions within their airport app or website to help passengers with trip planning. This could include the following:
- Detecting and predicting airport-specific issues that would arise if a passenger books a flight itinerary that involves more than one airline. This information would also greatly benefit an airport operator’s signage program. For example, would the passenger need to change terminals, and if so, does the airport have appropriate signage to guide them from one terminal to the next? This information could even be provided to travel booking sites, which may use their own algorithms to warn passengers or keep them from selecting connections that are too tight.
- Providing updated travel times to and from the airport.
- Providing on-time performance data for specific flights.
Some of these features are provided today by other entities, such as travel booking sites, but they could be integrated into an airport’s passenger experience program and customized to fit the needs of local passengers.
Operational Improvements
Resource Allocation
Airport operators should consider the impacts of AI on their existing or future resource management system (RMS) solutions. This could be accomplished by discussing with current providers or by including AI requirements in future requests for proposals. AI integration could help an RMS make predictive adjustments or recommendations based on real-time information. This would help take current rules-based RMS solutions to the next level. Airports could also add AI-based predictive analytics platforms to other operational systems to optimize resource allocation, including personnel deployment and energy consumption, by maximizing efficiency and reducing costs.
Generative AI Tools
There are a growing number of generative AI tools capable of understanding and producing natural language text. These tools, such as ChatGPT or Google Bard, can function as a virtual assistant by responding to user queries, providing information, and engaging in text-based conversations. A great way for people in various airport operator business functions (e.g., planning, executives, information technology, operations) to use these tools is for brainstorming “conversations.” For example, top generative AI tools could offer great insights on developing five-year planning frameworks, crafting job descriptions, or identifying future innovative technologies and summarizing their impacts.
In an interview for this article, airport operator Dave Wilson, former director of airport innovation for the Port of Seattle/Seattle-Tacoma International Airport, described how he used ChatGPT for these and other use cases. Wilson took it one step further and developed a website—AirportGPT—to help airport operator leaders learn about generative AI tools. Wilson’s exchanges with ChatGPT contain examples of specific prompts and full conversations from both an airport operator’s and passenger’s perspective to demonstrate the types of value ChatGPT can bring to airports.
Full Analytics Solutions
AI-integrated cameras are beginning to be used in the airport industry to monitor the refueling of planes, the unloading of baggage, safety concerns on the airfield, and many elements of the plane itself. The end goal is identifying and mitigating delays and cancellations by analyzing historical data, weather patterns, and other relevant factors to predict flight delays more accurately. This predictive data can be shared with airlines to help them plan better for possible delays and let their passengers know if there are any disruptions. Several airports and airlines have begun implementing these solutions, such as JFK International Air Terminal, Cincinnati/Northern Kentucky International Airport, and American Airlines.[4]
Concessions and Businesses
AI solutions could also help an airport’s concession or business departments. This could start with feeding data such as flight demographics, passenger tracking data, concession usage (by time of day and location), and concession complaints from passengers (e.g., bottlenecks, long lines, quality) into an AI algorithm. This could uncover detailed insights on the everyday realities of concession performance and passenger behavior, and it could reveal new, strategic locations for concessions; underperforming concession areas; or a lack of concession signage. Some bad passenger experiences with concessions could be prevented just by installing better, dynamic signage guiding passengers to an open concession area. Beyond this, airports could also consider dynamic pricing for concessionaires, based on time of year or other factors.
Autonomous Solutions on Ramps and Airfields
Autonomous solutions are quickly coming to airport ramps and airfields, and AI is foundational to their operation. These autonomous technologies use computer vision and machine learning to develop smarter paths for operation and collision avoidance, along with other features. These AI-integrated solutions will allow more ramp activity to be done remotely and autonomously.
Many autonomous ramp technologies are already being deployed in Asia and Europe. In the United States, several companies are working with airlines to deploy automated ramp equipment, such as bag carts and airplane tugs. These airlines are currently putting large investments into autonomous technologies for ramp equipment and vehicles.
Another use case is autonomous, AI-enabled snowplows and robots that cut grass. These robots can greatly limit the cost and hours of effort dedicated to maintaining runways and fields at the airport. One example is the Oslo Airport, which is working with a provider of autonomous snowplows for their runways.[5] Additionally, in 2019, Winnipeg Richardson International Airport became the first North American airport to deploy an autonomous plow, called Otto.[6] In 2021, Changi Airport trialed a group of small robots, similar to household sweeping robots, that cut grass along a preprogrammed path. The robots were solar powered and could return to a holding station to recharge.[7]
Airport executives will need to ensure that there is adequate infrastructure to accommodate these new solutions, especially with the rise in common use environments, which limit the amount of available infrastructure at airports. These technologies also require high speed, low latency network connections (such as those provided by private wireless networks) and sufficient charging stations. These and other impacts of autonomous technologies will need to be considered.
Security Monitoring and Threat Detection
Many airports are already using security systems that incorporate AI and machine learning technologies to enhance their detection, monitoring, and response capabilities. These systems help track abandoned baggage, monitor entry points, and scan baggage or persons. For example, AI-enhanced systems can analyze closed-circuit television camera feeds to identify suspicious behavior and alert security personnel to potential threats in real time. AI can also assist in analyzing x-ray scans for prohibited items in luggage. (For more detail on this topic, check out Enhancing Airport Security: Transformative Role of AI Across the Industry, an Applied Technology in Airports article.)
Conclusion
AI usage is expanding quickly, and many airports are already participating with new security systems, autonomous vehicles, and passenger chat features, among other technologies. AI technologies will likely continue to grow more advanced as airports access higher data speeds and AI models are improved. Airport operators may want to monitor the industry for new deployments and advancements that could help their operations. Continue to Part II for a summary of the potential challenges that airport operators should keep in mind.
Other recommended reading on this site:
Artificial Intelligence (Technology Focus article)
References
[1] Artificial Intelligence, Merriam-Webster, Accessed May 6, 2020, https://www.merriam-webster.com/dictionary/artificial%20intelligence.
[2] Delta’s Futuristic New Airport Screen Can Show Personalized Flight Info to Multiple People at Once — Here’s How It Works, Business Insider, Erb, Jordan Parker, July 7, 2022, https://www.businessinsider.com/delta-parallel-reality-board-detroit-displays-personalized-flight-information-2022-7.
[3] Now the Airport Info Booths Are Going Virtual, Baskas, Harriet, November 24, 2020, https://stuckattheairport.com/2020/11/24/now-the-airport-info-booths-are-going-virtual/.
[4] Delta’s Futuristic New Airport Screen, Erb, https://www.businessinsider.com/delta-parallel-reality-board-detroit-displays-personalized-flight-information-2022-7.
[5] Next Step in Autonomous Snow Removal at Norwegian Airports, Semcon, Accessed June 2023, https://news.cision.com/semcon/r/next-step-in-autonomous-snow-removal-at-norwegian-airports,c3321638#:~:text=Norwegian%20airport%20operator%20Avinor%20is,one%20of%20the%20co%2Downers.
[6] Winnipeg Airport Unveils Self-Driving Snowplow Named Otto, HuffPost, Shakeri, Sima, March 17, 2019, https://www.huffpost.com/archive/ca/entry/winnipeg-airport-unveils-self-driving-snowplow-named-otto_ca_5cd58f3fe4b07bc729793681.
[7] Robotic Grass Cutters Now In Use at Changi Airport, More to Be Deployed, The Straits Times, Wei, Toh Ting, November 22, 2021, https://www.straitstimes.com/singapore/transport/robotic-grass-cutters-now-in-use-at-changi-airport-more-to-be-deployed.
Header photo by Igor Omilaev on Unsplash, modified by Barich Inc.