AI Tools Take Aim at Airport On-Time Performance Challenges

• Why AI adoption is accelerating across airport operations

• How data quality directly impacts AI effectiveness and trust

• How airports are using AI to improve turn times, sustainability, and safety

Despite decades of technological progress, airport operations still face a persistent challenge: how to translate vast amounts of data into meaningful action. That was the central theme of a recent industry webinar hosted by Serium, ACI World, and Assaia, which brought together AI and aviation experts to discuss practical applications of artificial intelligence in improving on-time performance (OTP).

The discussion opened with a pointed reflection by Mike Malik of Serium, who recalled standing in a system operations control center 25 years ago. Even then, he said, the problem wasn’t the lack of data—but the inability to act on it quickly enough. Today, he noted, airports have access to more operational information than ever before, but many still struggle to move from reactive to proactive decision-making.

“Artificial intelligence is not some futuristic concept—it’s a practical tool that can help us predict disruptions before they happen, identify patterns we don’t see, and suggest solutions that human operators might not consider,” Malik said.

The Case for AI in Airport Operations

Kevin Hightower, chief innovation officer at Serium, emphasized that the pressure on airport teams is intensifying. With limited staff and growing demands, AI can provide a critical layer of support.

“Airport staff are wearing more hats than ever. AI can help scale operations by handling routine analysis and enabling people to focus on complex exceptions,” Hightower said. He also pointed to AI’s potential to capture and transfer tribal knowledge—a growing concern as experienced personnel retire from the industry.

Niha Shek, vice president of product at Serium, highlighted how generative AI is already being used to assist airlines and airports with performance insights. Serium recently launched a generative AI tool specifically built to analyze on-time performance data, helping users identify causes of delays, compare performance trends, and benchmark against top performers in minutes instead of days.

“Instead of relying on monthly reports and manual spreadsheet work, the system allows airport teams to ask detailed questions about performance by route, aircraft type, or turnaround time. And it’s only getting better,” Shek said.

Turnaround Time in the AI Spotlight

A focal point of the conversation was the role AI can play in optimizing aircraft turnaround time—a key factor in on-time performance.

Yan William Kea, head of sales and marketing at Assaia, explained how their AI-driven turnaround management platform uses computer vision and real-time video analytics to provide granular visibility into apron operations.

“Despite all the tech in airports, turnarounds look pretty much the same as they did 50 years ago—manual, fragmented, and largely undocumented,” Kea said. “Our solution creates and shares objective turnaround data, reducing delays and helping gate planners make better real-time decisions.”

He cited examples from U.S. airports where Assaia’s system helped shave off 2–6 minutes per turnaround by automating notifications and increasing coordination between gate operations and air traffic control. In Seattle, for instance, the platform enabled faster aircraft gate assignments by signaling when a gate became free, eliminating unnecessary wait times for inbound aircraft.

Data Quality: AI’s Backbone

Throughout the webinar, panelists repeatedly returned to the importance of high-quality, accessible data. While AI tools are improving rapidly, their usefulness hinges on the quality of inputs.

“Generative AI isn’t magic,” Hightower noted. “If you feed it bad data or leave it guessing, it can ‘hallucinate’—producing inaccurate or misleading results. The challenge isn’t the AI, it’s the data.”

He outlined how Serium’s AI engine is designed to keep proprietary data local and secure, with patented methods to avoid sending sensitive operational information to external large language models. The system leverages Serium’s extensive aviation data sets—including schedules, satellite tracking, and fleet information—to provide grounded and trustworthy outputs.

ACI World’s Diederik Mink added that many airports still struggle with fragmented and siloed data systems. “To enable effective AI use, airports need centralized data integration and clear governance frameworks,” Mink said. “That includes defining data ownership, ensuring quality standards, and protecting passenger privacy.”

AI’s Role in Sustainability and Safety

While efficiency is a major driver of AI adoption, panelists also explored how AI can contribute to sustainability and safety goals. From dynamic HVAC adjustments in terminals to real-time APU usage detection, AI is being deployed to reduce emissions and optimize energy use.

“AI-controlled climate systems can adjust temperature settings in real time based on passenger flow and weather conditions,” Mink said, citing examples from Brisbane and other airports. “That supports energy savings without compromising passenger comfort.”

Assaia’s system also monitors apron safety events such as missing safety vests or improperly positioned equipment—insights that can inform training and reduce incidents. And as Kea noted, the platform can even detect when an aircraft’s auxiliary power unit is running unnecessarily, helping airports and airlines reduce fuel burn and emissions.

The Roadblocks Ahead

Despite strong enthusiasm for AI adoption, webinar attendees identified integration with legacy systems as the biggest challenge, followed by internal expertise and regulatory hurdles. Panelists acknowledged that AI deployment requires cultural and organizational change—something that takes time.

“There’s no one-size-fits-all governance model,” Mink said. “But airports should treat data as a strategic asset, start with small wins, and build momentum from there.”

As global passenger volumes are expected to double by 2045, the demand for smarter, faster airport operations is only growing. AI offers a way forward—but only if it's built on quality data, thoughtful governance, and real operational alignment.

“AI won’t replace people,” Shek concluded, “but it can unlock capabilities we didn’t have before. And when it’s used to support—not replace—decision-makers, it becomes a real force multiplier.”