How GIS and AI Modernized San Bernardino Airport Airfield Inspections

San Bernardino International Airport (SBD) serves Southern California’s Inland Empire, supporting commercial cargo operations, general aviation, and a growing portfolio of passenger services. The airport manages more than 1,300 acres of infrastructure, including a 10,000-foot runway, taxiways, aprons, and safety areas that require continuous monitoring and maintenance.

Operating at this scale places constant pressure on airfield operations teams. Surface conditions directly affect safety, operational continuity, and regulatory readiness. Even small defects, if left undetected, can escalate into costly repairs or compliance risks. For SBD, maintaining a clear, up-to-date understanding of airfield conditions is mission critical.

To meet this challenge, SBD is exploring how airfield inspections can better support safety, compliance, and long-term infrastructure planning. By combining artificial intelligence (AI) with geographic information system (GIS) technology, the airport can capture inspection data at greater scale and consistency, creating a clearer picture of airfield conditions as they develop.

Integrated with ArcGIS software, this AI enabled inspection approach moves SBD beyond periodic, manually recorded observations and toward a more complete, operational view of the airfield. This shift strengthens day-to-day airfield operations while supporting Federal Aviation Administration (FAA) Part 139 compliance.
 

Part 139-certified airports must regularly inspect and document the condition of runways, taxiways, markings, lighting, and safety areas to meet FAA standards for commercial service. At SBD, trained personnel conduct these inspections through daily visual checks, driving the airfield to identify and record potential issues.

During each inspection, staff rely on personal experience and judgment to flag discrepancies such as pavement defects, foreign object debris (FOD), vegetation encroachment, and faded markings. This approach works well for identifying clear, immediate hazards, but it limits how much information inspectors can capture during a single pass.

A typical manual inspection produces approximately 20–40 human-logged observations per shift, often focused on the most obvious or severe issues. Results can vary by inspector, lighting and weather conditions, and time available. As a result, subtle early-stage pavement cracking, low-contrast marking wear, or small debris may go unnoticed or be recorded inconsistently.

Over time, these constraints create a larger operational challenge. Manual inspection records often live across paper logs, spreadsheets, and reports, making it difficult to understand how conditions change or compare trends across inspection cycles and locations. Questions about where deterioration is accelerating, whether repairs are holding, or which issues recur most frequently become harder to answer when inspection data remains fragmented and subjective.

To modernize how airfield conditions are captured and analyzed, SBD turned to Esri partner Airwai, Inc., which is developing AI-driven inspection capabilities designed to operate directly within the ArcGIS environment. By combining AI with location intelligence, the airport applies machine learning models to inspection data and organizes results spatially, creating a consistent foundation for understanding airfield conditions.
 

Airwai’s AI-powered inspection platform, Layered Autonomous Inference and Reasoning Agents (LAIRA), expands the scale of each inspection. During a single pass, the system can capture more than two million machine-ready data points across the airfield, describing surface features, visual patterns, and spatial context that manual inspections often miss.

Machine learning models then translate that raw data into structured inspection findings. The system automatically detects and classifies pavement defects, FOD, vegetation encroachment, and marking degradation, with outputs integrating directly into ArcGIS workflows rather than remaining isolated within a separate inspection system.

“Airports are entering a new era where decisions can no longer rely on periodic checks and fragmented data,” said Mike Burrows, CEO of San Bernardino International Airport. “By deploying AI-driven airfield intelligence within our GIS environment, we are demonstrating how airports can move from reactive operations to predictive, data-led management.”
 

Because the system applies consistent detection logic and severity scoring, inspection results no longer vary by inspector, shift, or conditions. That consistency allows SBD to compare results reliably across inspection cycles, laying the groundwork for trend analysis, predictive planning, and future automation.

In side-by-side comparisons, AI-based inspections identify three to seven times more surface anomalies per mile than manual inspections, particularly early-stage issues that are difficult for the human eye to detect reliably at operational speeds. This expanded coverage strengthens the data foundation SBD uses to manage airfield operations and make informed decisions.

Using ArcGIS Online, SBD staff turn AI-generated inspection detections into map layers to show where issues cluster, recur, or accelerate across the airfield. Heat maps and temporal filters help reveal patterns, such as recurring FOD hot spots or pavement sections that are deteriorating faster than expected.
 

Maintenance crews then use ArcGIS Field Maps to act on those priorities in the field. With precise locations already identified, crews move directly to complete repairs instead of spending time searching or revalidating findings. SBD estimates that this workflow could reduce field validation time by 50 percent or more. As crews update conditions from repairs, those changes that Field Maps updates synchronize in the airport’s GIS, ensuring that inspection results and field activity remain aligned.

ArcGIS Dashboards extends that field activity into a shared operational view. Airport leadership can see inspection activity and severity distribution across the entire airfield, helping coordinate teams and monitor emerging issues without relying on fragmented reports.

As inspection cycles repeat, the combined AI and ArcGIS workflow reveals how conditions behave, not just where they exist. By comparing results across inspection cycles, SBD can identify problem areas that recur, evaluate whether repairs are holding, and recognize early signs of deterioration before issues escalate into larger risks.

“From an operations standpoint, our priority is maintaining compliance while supporting the people doing the work,” explains Jonathan Galvan, manager at San Bernardino International Airport. “Our team has a consistent, objective baseline across shifts and conditions, strengthening our compliance posture without adding burden to staff.”
 

Over time, these accumulated inspection records form a practical digital twin of the airfield—a shared, GIS-based representation that carries operational knowledge forward. Rather than resetting with each inspection, insights remain accessible across teams and shifts, supporting daily operations, compliance reviews, and longer-term planning with a consistent, authoritative view of airfield conditions.

What began as an effort to improve inspections at SBD has evolved into a decision infrastructure built on AI-driven insights and ArcGIS, supporting safer operations, stronger compliance, and a more resilient airfield over time.