Enabling the Future of Aviation Safety AnalysisAugust 2018
"At MITRE, there's always somebody working to make the impossible possible," says systems engineer Hamza Syed. "I feel fortunate to be part of that culture."
After graduating from George Mason University in 2010 with a mathematics degree, Syed came to MITRE and immediately started tackling problems for which there is no obvious solution. And he wouldn't want it any other way.
"My department is focused on aviation safety, and my job is to discover new ways to take full advantage of MITRE's unrivaled collection of aviation safety data."
That data is the backbone of the Aviation Safety Information Analysis and Sharing initiative. A voluntary government/industry partnership, ASIAS collects and analyzes data from a wide variety of sources to advance aviation safety. MITRE is entrusted with the intake, protection, and analysis of that data.
By analyzing anonymized, or "de-identified," data from the FAA, airlines, aircraft manufacturers, individual pilots and air traffic controllers, and other sources, our researchers can proactively identify safety issues. They then provide that information to ASIAS's government and industry partners so they can address identified issues before serious incidents occur.
Using Big Data Analytic Capabilities Enhances Aviation Safety
This collection of ASIAS data is classic "big data"—datasets that are so large and complex that they are difficult to quickly and affordably capture, organize, and process with traditional systems and tools.
"The key to wrangling big data is having one operating system that knows about the thousands of machines used to collect and store various datasets and what data is located on each one of them," Syed explains. "When I first came to MITRE we had a new operating system that required the reconciliation of traditional mathematics—things I had learned in school—with things I'd never seen before. It required me to think in a different way to solve problems."
Never one to shy from a challenge, Syed quickly became adept at using the new technology and began using his expertise to merge the various aviation datasets that had been entrusted to MITRE, while protecting sensitive data. Clearly he's successful at what he does—he recently received recognition at the 2018 BEYA STEM Conference in the category of Most Promising Engineer.
Finding Creative Ways to Fuse Sensitive Data
That work formed the foundation for Data Fusion, a project with even greater impact. Data Fusion enhances ASIAS's ability to detect safety issues. As part of that effort, in 2013 MITRE asked ASIAS's member airlines to provide fully identified pilot safety reports, a departure from previous practices regarding this sensitive information. That meant the reports could be connected with specific flights. MITRE also asked the airlines' permission to re-identify flight telemetry data—detailed information about the positions of individual aircraft—using a novel track-matching algorithm Syed had created. With that algorithm, the telemetry data could be fused with weather and radar data to provide a more complete picture of what occurred on a flight.
But there was a catch. This was sensitive data, and the airlines would agree to these requests only if MITRE could meet some very stringent requirements.
Syed enabled MITRE to meet those requirements. He designed an architecture solution that protected the airlines' proprietary information while still allowing MITRE's safety analysts to fuse and learn from the additional data. That transformed ASIAS's ability to identify safety issues.
"Previously, we had weather and radar information, but we didn't have information about the state of the aircraft or anything the pilot might have said," he says. "So we didn't have a full picture."
"With Data Fusion, we get a 360-degree perspective on events that present a safety concern. We can also do this in the aggregate and analyze all instances of a certain type of event. For instance, did they have weather in common? Were the pilots executing a similar maneuver? Were there commonalities or a lack of commonalities? All that information helps us identify causal factors that can then be addressed."
Exploring the Advantages of Cloud Computing
In more recent work, Syed researched and designed a cloud-based architecture that could quickly and cost-effectively perform some of the computations used in non-sensitive ASIAS analyses.
"This is one of the great things about working at MITRE. I can pursue challenging problems that don't have clear-cut solutions. Having that latitude is one of the things I like best about the company. I've also worked with some brilliant folks here. This is a place where you can really get cross-domain pollination. As a result, we can accomplish things that wouldn't be possible in a siloed environment."
Syed also credits MITRE with supporting the advancement of his education. "While working full-time, I earned my graduate degree in systems engineering from the University of Virginia. MITRE gave me the flexibility to do that, and reimbursed a portion of my tuition."
To anyone considering working at MITRE, Syed says, "Don't be afraid to take risks. Calculated risk-taking is critical to finding novel solutions, and MITRE supports employees in doing that. There may be times when those explorations don't work out, but we learn from those experiences. And that can ultimately lead us to success."