Unlocking the Full Potential of Safety Data Through Trust and Collaboration

The final day of AP-SAS 2025 emphasised the power of shared safety data, the human role in identifying risks, and the importance of trust in reporting systems. Speakers highlighted the evolving use of AI and technology to turn raw data into actionable intelligence, while stressing that collaboration, leadership, and human insight remain central to advancing safety across the aviation ecosystem.

Freely and openly sharing information is how to truly unlock the value of safety data and human reporting, Thomas (Tass) Hudak, Site Leader, MITRE Asia Pacific Singapore, said in opening Day 3 of AP-SAS. In his remarks, Hudak outlined some of the origins of safety information sharing. A milestone event was TWA Flight 515, which crashed near Washington Dulles International Airport in 1974 at least partially as a result of miscommunication between the aircraft’s pilots and air traffic control. During the ensuing investigation, it was determined that another commercial flight had encountered the same situation on the same approach to the same airport just weeks before the accident flight. The Aviation Safety Reporting System (ASRS) was established within a couple of years of the accident, providing a confidential way to report safety issues without fear of punishment.

Other significant developments he cited included creation of the Commercial Aviation Safety Team, aviation safety action programs, and the Aviation Safety Information Analysis and Sharing program.

Hudak said that most safety issues are first identified by humans. “Every person is a sensor in the network.” When those observations are coupled with the data from automated monitoring programs and other technological advances, it enables the identification and analysis of possible precursors to accidents and incidents.

Collaborating within and across organisations further enhances safety, but as a number of speakers said during the three-day summit, trust is an essential element in any safety information reporting program.  He said it takes patience and persistence to earn trust, build partnerships, and break down barriers to collaboration.

“With trust, strong leadership, and collaboration, we can keep pushing the envelope to make aviation safer,” he said.

Despite advances in reporting, data collection and analysis, and industry collaboration, there are still blinds spots in what isn’t reported. It’s still a challenge to find the unknown unknowns, he said.

Hudak added that investments in technology are needed to advance the state of the art in vulnerability detection to rapidly and efficiently identify and mitigate risk in a fast-changing environment.

The intent of any safety intelligence function is to turn raw data into actionable insights, but reaching the point where data is returning maximum value to an organisation may take time and multiple steps, according to speakers during Thursday morning’s Power of Safety Data panel discussion.

John Thomson, Senior Technical Advisor–Safety Management, United Kingdom Civil Aviation Authority International, who moderated the discussion, said the first round of analysis may result in a simple graph that shows increasing or decreasing numbers, trends or patterns, depending on the issue. The next step of analysis may produce a judgment using knowledge and experience and often results in a request for more information or data for a more thorough investigation. The final level is when the analysis produces a validation or more thorough understanding of what is going on, which can lead to recommendations or actions.

Thomson stressed the need to protect the individual making the original report, as did other members of the panel. Chantal Berthiaume, Director Business Systems and Performance, International Air Transport Association (IATA), said, “We don’t even let our own analysts know” where the data comes from.

Artificial intelligence (AI) can be used in a variety of ways to analyse the data. Berthiaume said IATA accepts accident and incident reports in more than 20 languages and uses AI to identify the language and translate the information. She also said that different regions have different reporting cultures and processes, so IATA analysts work to identify and understand biases that may exist in the data.

Shawn Lee, Head of Digital Solutions, Airbus APAC, said technology is needed to translate data into actionable intelligence, but he and others stressed that humans need to be in the loop. He also said industry needs to balance the use of AI and other processing-heavy tools with their environmental impact.

Oliver Chamberlain, Vice President Manufacturing, Rolls-Royce Singapore, stressed the need to create the right “leadership shadow.” For instance, he said he walks the floor of their facility every day as a way to encourage personnel to speak up. He also said it is important to have the right reporting tools in place and to make sure employees know how to use them. Ease of reporting is an important consideration, he said.

Chris Ranganathan, Chief Learning Officer, CAE, said training data is often not considered safety data, but he pointed out that training sessions in a simulator are often the only way to collect data on how flight crews perform in rare or extreme situations. He said it is a wasted opportunity not to incorporate training data into a safety management system.

In addition to sharing safety intelligence, Berthiaume urged attendees to share expertise. If your organisation is good at data analysis, “you should help others on that path,” she said.

Thank you to all our AP-SAS supporters