AI on the Tarmac

EVOLVE 2026 · MIAMI · PANEL · SESSION 06

Airport executives from Atlanta and Dallas-Fort Worth, and a wildlife-AI founder, discuss what it actually takes to deploy AI safely inside some of the world’s most complex operational environments.

EVOLVE is Sand’s annual summit bringing together technology leaders, policymakers and practitioners working at the intersection of AI and critical infrastructure.

Airports are among the most operationally complex environments in the world. Every day they coordinate aircraft, passengers, weather, wildlife, security, infrastructure and thousands of frontline staff—all while maintaining an uncompromising standard of safety.

This session explored what it really takes to introduce AI into that environment. The discussion quickly moved beyond technology to the operational realities of procurement, regulation, partnerships and organizational change.

Watch The Full Session

THE 60-SECOND READ

The panel challenged a common assumption about AI adoption: technology isn’t the hardest part. Operational complexity is.

Whether discussing wildlife management, winter operations or passenger experience, the speakers agreed that deploying AI inside critical infrastructure requires far more than good models. It requires trusted partnerships, procurement processes that can keep pace with innovation and solutions that integrate seamlessly into existing operations.

What This Conversation Revealed

1. Critical infrastructure adopts AI differently.

Airports cannot experiment with the same tolerance for failure as many other industries. New technologies must prove they can improve operations without compromising safety, reliability or regulatory compliance.

2. AI succeeds when it fits existing operations.

Across every example—from wildlife monitoring to winter preparedness—the discussion reinforced the importance of integrating AI into established workflows rather than asking operators to replace systems that already work.

3. Solving one bottleneck often exposes another.

Whether managing passenger flow, baggage handling or airfield operations, the panel repeatedly highlighted that optimizing one part of a complex system is rarely enough. Real transformation requires understanding how every operational component connects to the next.

Key Highlights

Preparation matters as much as prediction. Justina Mann recalled a winter storm where an airport had to truck salt in from Wisconsin because none had been stockpiled locally. The lesson wasn't about weather forecasting—it was about operational readiness.
The cost of disruption is enormous. A single snowstorm cost Delta an estimated $60 million, while specialized snow equipment costing $3–4 million per unit may sit idle for years between major weather events.
Safety remains non-negotiable. Ricky Smith explained that even seemingly small delays in aircraft de-icing can trigger cascading operational impacts, requiring constant coordination between air traffic control and ground operations.
AI augments frontline expertise. Flox Intelligence's wildlife monitoring systems are designed to support—not replace—human wildlife teams, helping them focus attention where risks are greatest while reducing manual patrols.
Modernization is a systems challenge. Atlanta is introducing facial recognition to streamline passenger processing, but Ricky Smith noted that expanding gates without also upgrading supporting infrastructure simply shifts congestion elsewhere.

By the Numbers

$ 0 M

Delta's loss from a single Atlanta snowstorm

current Atlanta airport staff

108M → 160M 0

ATL passengers a year, current vs. planned

0 ft

range of a single wildlife sensor

In their words

“Rule number one: know your customer. Times ten.”

Justina Mann, former VP, Dallas-Fort Worth International Airport

Why it Matters

Airports offer one of the clearest examples of why deploying AI is fundamentally different from demonstrating it.

Unlike controlled pilots or standalone applications, airports operate as tightly connected systems where every decision affects safety, efficiency and passenger experience. Introducing AI therefore isn’t simply a technical exercise—it requires navigating procurement, regulation, operational workflows and organizational trust, often simultaneously.

The discussion also highlighted an important lesson for Physical AI more broadly: intelligent systems create value when they fit naturally into existing operations. The examples shared—from wildlife monitoring to winter preparedness—showed AI succeeding not by replacing experienced operators, but by helping them anticipate problems, allocate resources more effectively and make faster decisions under pressure.

Perhaps the session’s biggest insight was that modernization is rarely achieved one technology at a time. Complex systems evolve when organizations improve the connections between people, processes and infrastructure. AI becomes most valuable not when it optimizes a single task, but when it helps the entire system perform better.