There is a lot of talk about data sharing as the next big safety challenge, with good reason: Data sharing is the key to achieving the next level of safety, and data sharing is really hard to do.
Accidents often have their roots in fairly obscure events. The most recent example is the crash of Air France 447. The pitot-tube failures and airspeed anomalies that initiated that tragedy were so rare that a single airline, even a big one, would not have been able to amass enough data to see it coming. Yet after the accident, when everyone shared their information, a dozen or so similar cases jumped out. We have been striving for proactive safety management for nearly two decades, but our data are still walled off. We can never really predict the next failure until those walls are broken down.
So why don’t we just all get together and do it? Probably because it is a lot harder than it looks. It is tough sharing data between just one airline and its regulator. In many countries, regulators are required by law to prosecute any violation of which they become aware. That discourages an airline from handing over all of its information. Even so, regulators often end up with more information than they could ever analyze, and if they did dig through all of the information, they would find a lot of things they would rather not know and would rather not act upon. To really use the data, the regulators have to be blessed with great technology that lets them glean insights from the mass of data, and then be allowed some discretion as to how they will act. In the real world, both of those advantages are in short supply.
The United States has fared better than most in this regard. It has amazing technology called ASIAS (the Aviation Safety Information Analysis and Sharing system) that lets it fuse voluntary reports, flight data monitoring (FDM) data, weather, air traffic control information and other data into one comprehensive picture. The United States also has special legislation that lets regulators accept these data from airlines without being compelled to act on every possible mistake. This magic combination has allowed 43 airlines to step forward and share a stunning amount of data.
Unfortunately you cannot just bottle the U.S. experience and export it elsewhere. Other counties don’t have a large enough aviation system to justify such a big investment in analytics. They would have to pool their data with neighboring countries. It is hard to develop appropriate trust between one airline and one regulator. Imagine developing that kind of trust between dozens of regulators and airlines that do not necessarily get along. Data sharing is vital, but in the real world it is not a turnkey proposition.
This is an area where the Foundation is working today, and it will be part of our focus for the foreseeable future. We can help regions find or develop information-sharing technologies. But more importantly, we also are in the position with regulators to develop the delicate arrangements that will allow them to share their data across borders, and then act responsibly on the data that the industry has entrusted to them.