Why Airports Need Machine Learning for Passenger Flow
Part of Xovis’ Airport Summit series, the two-day community event covered talks many topics, including how Artificial Intelligence (AI) and Machine Learning (ML) can be used to improve passenger flow and better manage costs.
AI and ML are incredibly broad concepts generally understood as computers’ abilities to perform certain tasks at a level equal to or greater than humans. In the passenger flow management segment, AI and ML are most related to visual detection capabilities like accurately identifying people in groups, such as -queues, and self-learning and self-adjustment without human intervention.
As in other industries, AI and ML are so important because they can dramatically increase productivity and operational efficiency without drastically increasing operating or staffing costs.
Using self-learning solutions to reduce passenger wait times
Xovis’ Airport Summit US took place at Newark International Airport (EWR) new Terminal A amid revolutionary changes in how businesses use AI and ML. The conversations in EWR meeting rooms, with leaders from some of the biggest and best airports in the Americas, covered many of the same questions being asked in other industries where mission-critical processes can benefit from AI-powered optimization.
Maximizing space utilization and avoiding resource overload are the main goals of passenger flow management, which is, at a fundamental level, about making real-time adjustments and iterative improvements based on accurate measurements of passengers in a designated space. The resulting benefits are many and include lower staffing costs, happier passengers, and higher land and airside revenue.
Modern queue detection and measurement solutions are automated, but very few have self-learning and self-adapting capabilities. These features—developed with the most advanced ML algorithms—are one of the areas where AI will deliver the most value to airport operators that rely on passenger flow management to bridge service gaps.
Navigating the AI Frontier
Many businesses, including airports, sometimes learn the hard way that tech disconnected from real-life situations is a high risk. Concerns about unfulfilled promises are understandably high in the current environment, where it can be difficult to separate function from fantasy.
Coverage offered by technologies such as LiDAR and traditional cameras is defined by physical principles, with occlusion determining the accuracy of measurement in crowded spaces. But the effectiveness of coverage can only be gauged in real-world experiences, gained from actual work at airports.
Both LiDAR and CCTV have some AI capabilities but are more limited in how further advances can be incorporated effectively and timely. And while LiDAR doesn’t have the same privacy issues that limit how CCTV is used in many jurisdictions, the underlying technology is prone to light-based distortions and has a notoriously short useful life.
Airports globally use Xovis’ Passenger Flow Management System (PFMS) because the sensors underlying the solution combine 3D information, like that provided by LiDAR, and full vision images, like those delivered by CCTV systems. That AI features can be added to sensors, allowing for advanced image analysis and processing, is another reason operators choose our future-proof solution. In addition to respecting privacy rules, Xovis sensors also boast a 25-year Mean Time Between Failure.
Of equal importance for airport operators is Xovis’ strong reputation for delivering robust, market-proved solutions. We have been helping airports solve overcrowding and queuing problems for more than a decade, acquiring specialized knowledge that can only be learned in practical situations. New and existing customers benefit from the best practices we’ve developed helping more than 120 airports solve real-world challenges, know-how that can’t be synthesized.
A Commitment to Innovation Partners Count On
Xovis was a pioneer in sensor-based passenger flow management and has never strayed from its position as an innovative disruptor. We continue to invest in AI and ML applications and solutions that solve resource overload challenges in airports, regardless of size or queue complexity.
In Newark, an attendee mentioned how Xovis sensors installed at her airport 10 years ago were recently updated with our latest machine learning-based algorithms. For me, this is an excellent demonstration of Xovis’ future-proof approach: long-lasting quality designed to constantly evolve with new advances.
AERO, our new fully managed service, reflects that approach. The cloud environment provides a secure operating framework and guarantees consistent and reliable delivery of accurate data. A cloud-native solution spares operators from the cost and inconvenience of downtimes associated with on-prem solutions and can also eliminate the need for specialized staff.
Tags: | airports| AI airports | passenger experience | data quality | terminal operations |
