Technology Description
The business and technology community has long anticipated the benefits and potentially drastic disruption brought by artificial intelligence (AI). Understanding the reach and impact of AI has been an issue amongst researchers for decades. Ambiguity around AI has caused numerous fears and assumptions about the impact on jobs, security, privacy, and other areas of life yet to be identified. While AI has not yet materialized into the uncontrollably disruptive force that some have predicted, impacts are already being felt in certain domains due to recent advancements in AI functionality. Industries have been looking to make sense of how to use AI within their organizations as well as how to prepare their organizations for the pending impacts AI solutions may bring.
The specific definition of AI can vary, but the Merriam-Webster dictionary defines “artificial intelligence” as “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]
The distinction must also be made between other technologies commonly used interchangeably with AI such as machine learning and deep learning. Machine learning is a subset of AI, and it consists of the techniques that enable computers to identify trends and begin to develop an understanding of certain topics. Deep learning is a subset of machine learning that enables computers to solve more complex problems. Simply put, each of these are aspects of AI, the main difference being in the level of complexity, with deep learning being the most sophisticated.[2]
AI can be further defined by the study of “intelligent agents,” which are any device that perceives its environment and takes actions that maximize its chance of successfully achieving its goals. This is why the term “artificial intelligence” is often used to describe machines (or computers) that mimic “cognitive” functions that humans associate with the human mind, such as “learning” and “problem-solving.” However, one of the benefits of AI is its capacity for seeing what is happening in ways that humans simply cannot.
Since airports are both producers and consumers of vast amounts of interconnected data, and since that data can provide airport organizations with deeper insight into airport performance, it is important that airport operators have a way to leverage this vast amount of data in real time. AI can help with this by providing ways to identify trends and make data-driven decisions at speeds and levels of efficacy that are not achievable with human analysts alone. Therefore, to effectively manage large organizations while still driving efficiency, airport operators can begin to incorporate AI technologies.
AI was placed into the Advanced Transformation Tier of this Publication due to the potential wide-reaching impacts the technology may bring to airport operations. Although forms of AI are leveraged within airport organizations today, the advancement of the technology could bring enhanced automation functionality to processes impacting travelers and airport staff. Although AI may bring numerous performance gains to airports, airport operators must be prepared to address AI and automation’s disruptive effects on the airport industry. AI’s ability to impact consumers and staff through the introduction or elimination of processes makes the technology a priority for airport operators to monitor.
Impacts
Until real-world deployments begin, airport operators will likely not understand the full impact that artificial intelligence (AI) deployments will have on their airport environment. This article outlines many of the high-profile impacts that airport operators may expect from the Internet of things devices as their usage increases. Following the impact list, two notable impacts are detailed further.
Impacts
Management/Operations
- Cost reduction through efficiencies identified by AI solutions
- Reduction in staffing cost due to the inclusion of automated solutions
Technical/Infrastructure Readiness
- Data integration necessary for AI solutions to identify trends and provide solutions
- Updates required to current computing infrastructure to support the needs of an AI solution
- The proliferation of connected equipment and processes to gain data from current airport operations
- Software solution redevelopment to leverage AI and machine learning functionality
Process/Skill Set Changes
- Airport processes must incorporate insights gained from AI
- Replacement or augmentation of human staff with automated elements
Passenger Experience/Passenger Process
- On-demand passenger service across the whole passenger journey through automated communications platforms
- Passenger trust in AI solutions is necessary
- Staff gain capacity to help passengers with more specialized tasks
- Passenger trends can be identified to improve the passenger process
Security/Safety
- AI solutions can identify security threats across the airport property at a higher rate than solely human personnel
- Increased cybersecurity threat as AI-enabled malicious software can more efficiently exploit system flaws
Airport Design/Construction
- Airport designers can leverage insights gained from operational data to design solutions that improve the performance of airport processes
Revenue/Business Model
- Increasing revenue through targeted advertisements based on passenger segments
- Using passenger trend data for tenant choice and concession location along with product choice to increase customer spend
- Leveraging automation to improve aircraft turnaround times
Legal/Risk
- Data ownership rights, like the owners of data sources, can vary
- Public data privacy laws—operators must consider regulation in using third-party data for airport use
- Previous labor contract considerations
Featured Impacts
Targeted advertisements based on passenger segments: Airport organizations can leverage passenger segment trends to improve the effectiveness of advertising. Using methods similar to online marketers, airport marketing departments can leverage insights that provide contextual and targeted ads based on passenger behavior. Improved advertising data can increase the price that ad providers are willing to pay while increasing revenue through alerting passengers to goods and services in which they are interested.
Previous labor contract considerations: Incorporating automation and AI into airport processes may identify labor redundancies or the lack of need for certain staff. However, replacing the current staff with an automated solution may not be simple due to existing labor contracts. The stipulations outlined for current contracts must be understood to confirm that a reduction in staff numbers or efforts will not violate these agreements. Airport operators negotiating future labor contracts must consider technology changes to not be locked into a disadvantageous situation for the airport business. With that said, it should not be assumed that jobs for these staff simply disappear. This is what Changi Airport experienced following the deployment of biometric solutions, which increased opportunities for passenger self-service options. Airport staff were transitioned from passenger assisting roles and retrained to monitor the automated solutions.[3]
Attributes
To better understand artificial intelligence (AI), airport operators should gain a better understanding of its specific attributes. This article explores the operating factors of AI and some of its usage characteristics. Understanding these attributes will help airport operators determine applicable use cases for AI and how their organization can support it.
Elimination of Current Processes or Approaches
Automation can be applied to processes across the airport environment to improve performance and efficiency. AI automated solutions can limit the need for human involvement in a process or even eliminate whole processes. However, airport organizations can invest in the reskilling of airport staff to perform other, much-needed jobs. These can include roles offering more personal customer service or undergoing retraining for roles to assist with the management of automated solutions.
Computing Power, Hardware Requirements
AI solutions require extensive computing power to keep up with the vast amounts of data AI can process. Effective software solutions must also be developed based on the planned use cases.
Learning Curve
Airport organizations will need to contend with a potentially steep learning curve when deploying AI solutions within the organization. This includes ensuring that their staff has the skill sets to program and maintain AI solutions and understand the best use cases for AI.
Use Cases/Business Effect
With every new or emerging technology, there are two basic questions an airport operator asks: “How can my airport use this?” and “How does this affect my business, even if my airport doesn’t wish to use it for our own benefit?” This article provides answers to both of these questions, addressing the airport uses cases and business effect of artificial intelligence (AI).
Use Cases
As AI solutions advance to a point of being cost-effective with widespread adoption, potential use cases become available for airport operators to incorporate AI into their organizations.
Trend Analysis and Decision Making
AI provides airport operators with an unprecedented ability to ingest the abundance of data presented at an airport location and identify data trends that can influence decision making. Airport organizations can monitor operational performance, forecast revenue, and locate factors that can positively or negatively influence performance.
Customer Service
Seeking to reduce passenger stress is an important factor for airport operations. One notable area of stress for many passengers is airport navigation. AI solutions can provide on-demand service to passengers in need of assistance. Sophisticated solutions can address passenger issues and provide potential solutions, such as with Munich Airport’s Josie Pepper, a humanoid customer service robot equipped with AI.[4]
Business Effect
Whether or not an airport chooses to employ AI for its own benefit, AI deployments that are not directly initiated/driven by the airport organization may have an effect on aspects of the airport’s operations.
Provide Access to Data Sources
As AI technology relies heavily on data, external organizations leveraging AI solutions will request access to more data sources around the airport to draw their own insights. Airport operators must design policies around what data sources are available and design the proper data integrations for these external parties to have access.
Biases in AI Solutions
Systems designed using poor quality reference data may exhibit errors in the system’s decision-making logic, an issue referred to as “system bias.” Bias appears in solutions that have used flawed data sets, where the flaws in the data become evident in the solution’s performance. Biased systems may not operate as intended when faced with certain segments of users or certain environmental scenarios. Airport organizations must properly vet companies looking to incorporate their own AI solutions into the airport to eliminate systems with flawed logic. AI solutions with operating biases can negatively impact customer service and reduce user trust in other AI solutions.[5]
Tiered Approach
Airport operators interested in new or emerging technologies, such as artificial intelligence (AI), will differ in their levels of risk tolerance. Some organizations are comfortable at the forefront of technology and have the resources to support innovation. Other organizations are interested in simply exploring how they can use AI within their limited resources.
This article takes a tiered approach to AI technology, providing use cases that are separated by the following innovation tiers: Reactive, Strategic, and Innovative.
Reactive
Artificially Intelligent Chatbot
Airports can employ off-the-shelf chatbot technology to engage with customers in need of quick information. Passengers can simply ask their questions to the AI chat client and receive instant answers to most questions without the need for human interaction. Similar solutions have been deployed in both aviation and non-aviation industries and help to free up customer service staff for more specialized requests while providing customers who have easier questions with instant service.
Strategic
Automated Smart Building Solutions
Buildings with network-connected infrastructures—such as lighting or heating, ventilating, and air conditioning systems—combined with AI software can dynamically adjust infrastructure to changing environmental factors, providing cost and operational benefits.
Innovative
Automated Airport Resource Management Solutions
Airport organizations can deploy AI solutions that go beyond building infrastructure control, supporting decision-makers in areas such as staffing, dynamic pricing for commercial and passenger services, and potential processing improvement opportunities for airport operators.
Industry Status
Exploring artificial intelligence (AI) technology deployments in both aviation and non-aviation industries can provide airport operators with a better understanding of the technology as a whole. This article outlines the current state of AI technology from both perspectives.
Aviation Industry
Airfield Apron AI
An efficient airplane turnaround process allows airports to make better use of increasingly constrained resources. New solutions are using AI to monitor boarding gate turnaround performance.
Passenger Volume Forecasting
Airports can use AI to assess external factors and determine passenger volume. As systems become more sophisticated, they provide more detail and accuracy around passenger travel habits. AI can take information about passenger travel habits and identify trends that can be leveraged by airport operators to make the proper decisions.[6]
Customer Service Chatbots
Airlines and airports are deploying AI for customer service to help passengers with routine questions. This provides a quicker level of service during high volume times. Costs can be saved on staff while still providing high levels of service.[7]
Non-aviation
Digital Personal Assistants
Google, Amazon, and Apple have been leading the industry in the development of digital personal assistance. These solutions are able to answer questions and perform tasks for their owners.[8]
Targeted Digital Advertising
Algorithms trained on customer data can begin to identify market segments across millions of Internet users. These algorithms can then begin to serve up ads that are most related to the individual user, with the hope that these will have improved performance over less targeted advertisements.[9]
Technology Interaction
Technology solutions may enable or be supported by other types of technologies. In some cases, the advancement of one technology may be vital to the effective use of another. This article highlights some of the high-level ways that artificial intelligence (AI) technology may leverage the functionality of other technologies or be used to enhance the functionality of another technology. As solution development continues, integrations with other technologies may become more evident.
Internet of Things
AI solutions can identify correlations and draw insight from the data provided in connected devices. AI can help automate the management of these devices and be used to make real-time adjustments based on environmental factors.
Digital Twin
As AI becomes more advanced, it can facilitate more sophisticated digital twin operations. AI can process and forecast data based on simulated changes to equipment and processes. Advanced AI solutions allow airport operators to draw new levels of operational insight on airport processes and equipment.
Robotics
AI is a central component to more advanced robotics solutions, enabling robots to operate in varying use cases and environments. As robots are given more complex tasks, AI integration is required to make decisions as factors in the environment change. Robotic solutions with advanced AI can be deployed in sensitive and passenger-facing environments, allowing these robots to have a better awareness of their surroundings and make adjustments to unforeseen process changes.
Technology Barriers
As with other technologies noted in this Publication, there are market forces or obstacles in development that present barriers to the deployment and widespread adoption of artificial intelligence (AI) technologies. Keeping abreast of these barriers can help airport operators know when to expect to see wider use of this technology in the public, and when they should look to reassess it for their own use. This article outlines the current state of AI barriers.
Computing Requirements and Limited Functionality
Current iterations of AI solutions require heavy computing requirements and well-designed software to operate. Along with the computing hardware necessary, AI solutions can have difficulty adjusting to variable-rich tasks. These constraints mean that potential use cases for AI are relatively limited. However, as advancement in AI continues, new applicable use cases can be identified.
Data Integration
AI solutions must have proper data integrations to provide useful insight. The strength of an AI solution is its ability to analyze and draw correlations from big data sets. However, gaining access to these data sources can be difficult, especially due to the increasing value of data overall.
A Better Understanding of Use Cases for the Airport Environment
Airport operators should understand the best use of AI solutions to be successful within the organizations. AI advancement can lower costs and technical requirements needed to deploy and manage solutions. As costs are reduced, it can be easier to make a business case for AI within an airport organization.