Aviation Experimentation: Moving MITRE's Labs to the CloudOctober 2014
Topics: Aviation Industry, Airports, Airspace, Modeling and Simulation, Human Computer Interaction, Human Factors Engineering
In the last year, MITRE has made great strides toward applying cloud-computing technology to our aviation laboratory experiments. This advance both achieves time and cost savings for our sponsors and expands the number and variety of stakeholders who can take part in these experiments.
Our research on the potential of cloud computing as an alternative means to deliver laboratory experiments started in 2013. Sheng Liu, a senior engineer in MITRE's Center for Advanced Aviation System Development (CAASD)—the FFRDC sponsored by the Federal Aviation Administration—began working on a project to expand the reach of our human-in-the-loop (HITL) simulation capabilities beyond MITRE's walls by making them available in the commercial cloud.
"The idea was to support HITL experiments and demonstrations without physical space and resource constraints," he says. "We want to make our simulation assets more accessible. That means being able to run and share aviation tools and simulations with sponsors—wherever they're located—using cloud technologies."
In addition to exploring how to use cloud technology effectively, Liu assembled a team to investigate the myriad issues that affect an organization's ability to use this emerging capability, including information security, licensing, and performance issues. While initially focused on simulations in MITRE's Integration Demonstration and Experimentation for Aeronautics (IDEA) Lab, Liu's work is having an effect on experimentation across MITRE.
In the aviation domain, several cloud-related projects have spun off from Liu's work.
Testing a New Air Traffic Control Curriculum
The first of these was a project to create and test a curriculum that would provide future air traffic controllers with a foundation in traffic flow management. That's a skill becoming increasingly important as Next Generation Air Transportation System (NextGen) capabilities and procedures continue to be implemented.
The project's principal investigator, Roberta Zimmerman, began piloting this curriculum two years ago with students enrolled in air traffic control programs at Embry-Riddle Aeronautical University and the Community College of Beaver County in Monaca, Pennsylvania.
"For the first class, before we had access to the cloud, we had to physically recreate analogous lab facilities at the university, which was very time consuming," Zimmerman recalls. Upon learning about the work Liu's team was doing, that changed. "We collaborated with Sheng's team to deliver those lab capabilities to the students in their own classrooms, using their own computers, via Amazon's cloud services. It worked very well, and we'll continue to use this approach in future classes."
When that application of cloud technology showed promise, CAASD enterprise architect Brian Simmons established an initiative to develop more demonstrations of how to use cloud technology in ways that would provide direct benefits to the FAA.
"We secured internal funding for two use cases we believed would resonate with different MITRE sponsors," Simmons explains. One dealt with border security and the other with airspace design.
Broadening Participation in Airspace Design Testing
Several years ago, CAASD and the FAA embarked on a project to identify near-term changes that could reduce delays, increase throughput, and improve fuel efficiency at the nation's busiest airports. Specifically, this work focused on the air traffic in metroplexes, or regions served by multiple major airports.
New arrival and departure procedures have already been implemented in the Houston and Washington, D.C. metroplexes. By 2018, similar work will have been completed at 13 major metroplex sites. All of the designs developed for the metroplexes will be tested prior to implementation.
Until recently, all testing occurred in the IDEA Lab or via portable equipment shipped to FAA facilities. This either required controllers familiar with the airspace to travel to the MITRE facilities in McLean, Virginia, or required MITRE staff to set up and staff portable lab capabilities at FAA field facilities. Earlier this year, however, some HITL evaluations of the new designs were demonstrated via the cloud.
"Those demonstrations were very successful," says Airspace and Procedure Design and Analysis department head John Brandt. "This means FAA personnel will often not have to travel to MITRE for testing of our designs. In addition, because this new capability eliminates the requirement to set up a portable lab in the field, it enables a broader population of stakeholders to participate in these simulations."
Creating Multiagency National Airspace Security Scenarios
A joint effort between two MITRE FFRDCs—CAASD and the National Security Engineering Center (NSEC)—is also exploring the potential of cloud technology to achieve a more rapid and coordinated response to unauthorized air incursions coming into our national airspace
"When an unauthorized aircraft crosses into U.S. airspace, the response can require coordination among a number of different individuals and organizations in a variety of locations," says Simmons. "Our sponsor wanted us to set up a scenario to reflect those complexities and provide a mechanism that would facilitate the necessary coordination."
In particular, the sponsor asked that the scenario reflect the FAA's participation in responding to the threat. To accomplish that, the team took advantage of an IDEA Lab capability that virtualizes an air traffic control facility's data and response. That was combined with other MITRE capabilities to create an environment in which en route and terminal air traffic controllers could share some subsets of information with homeland defense and homeland security agents in various locations—in real time.
"We wanted to be able to put capabilities, data, and procedures in the cloud so that we could share them with field agents in their own settings, using their own equipment, whether that's a laptop or a smartphone," says National Security Experimentation Laboratory project leader Jim Dear, who works within NSEC. "This study helped us demonstrate the potential of the cloud to enable us to do that." The study is now informing future work with across multiple MITRE sponsors that will employ the cloud.
Making the Case for the Cloud
The success of these initial projects has also prompted CAASD to pursue other use cases for its FAA sponsors.
"We want to make as many sponsors as possible aware of the benefits of cloud technology and our ability to make use of it in pursuit of their objectives," says Simmons. "Use cases are the best way to do that. The cloud allows us to broaden our simulation reach and have a much richer, more robust simulation at essentially no additional cost.
"This technology is transforming the way our sponsors make decisions and the ways in which they use our services. It's a potential game-changer for how we can bring impact with our labs."
—by Marlis McCollum