How data and analytics enhances safety decision-making

The Safety and Technical Training service (AJI) is a branch of the Federal Aviation Administration’s (FAA) Air Traffic Organization (ATO) that is responsible for proactively identifying safety trends and mitigating risk using data-driven strategies. Additionally, AJI provides technical training for air traffic controllers and technical operations technicians and engineers so that they have the data points and insights to support their day-to-day activities.

The FAA collects and owns a great amount of data related to air traffic, aircraft and airmen certification, safety performance, etc.  For all of the data collected, the FAA recognizes that the data could be more useful, reliable and trusted. Continuing their multiyear journey, AJI leaders in FY23 looked toward data and analytics to deliver objective safety insights for the air traffic controllers and technical operations technicians and engineers that they support. AJI asked Ernst & Young LLP to build on work from past years’ collaborations to develop a solution that will advance the AJI agenda to provide tools and services that support risk-based decision-making for AJI’s customers.

The EY team of data scientists, data engineers and BI developers helped AJI create its first end-to-end data product that delivered value by integrating and analyzing the vast stores of data across a multitude of domains (safety, weather, equipment, operations, etc.). The safety insights enhanced by contextual data are visualized in a dashboard that is published and accessible to all FAA employees, providing clean, curated and integrated data that significantly decreases the effort required to identify safety trends from weeks to seconds.

We closely collaborated with air traffic controllers and technical operations personnel to better understand the safety trends and identify key indicators of risk in the National Airspace System (NAS). To do so, we conducted interviews to assess data needs and how they assess data quality and applicability. Those insights fueled the design of our advanced analytics and dashboard wireframes for iterative enhancement. The machine learning models were used to enrich the data and provide insights on safety trends.

Key benefits for AJI:

• Expanded analytics footprint across the ATO
• Integrated technical operations technicians and air traffic controller perspectives in supporting safety risk metrics development
• Provided a robust tool to identify safety trends that support safety discussions between technical operations technicians and air traffic controllers
• Streamlined the creation of quality assurance reports with data-informed safety trending
• Leveled up AJI’s overall data dexterity and knowledge of data science

In the end, an analytic-enabled dashboard was developed, providing drilldown analysis features, trends visualization and exportable data sets on high-quality, curated data. As a result, we demonstrated how AJI could leverage machine learning, natural language processing and advanced analytics as tools to improve data quality and improved data-informed decision-making.

Collaboration and data sharing was critical to the success of this engagement. We helped with data integration and interoperability to expand the view on safety and risk relating to air traffic. Users wanted a unified view of data and insights to build a more holistic awareness of each other’s work. Through our work, we were able to provide visibility and alignment across the organizations in the ATO to facilitate safety discussions.

Sharing information is an important benefit to AJI as air traffic and technical operations personnel possess invaluable knowledge that can be leveraged for safety trends and risk measurement. By facilitating these safety discussions, AJI can implement a beneficial feedback loop where data can be leveraged to validate or adjust existing safety hypotheses and drive toward leading indicators of risk and predictive analytics.

Going forward, we plan to work with AJI to further enhance its technical data capabilities through collaboration with their chief data office to build an enterprise data platform to manage safety data and scale machine learning solutions.

The views reflected in this article are those of the author and do not necessarily reflect the views of Ernst & Young LLP or other members of the global EY organization.