An ongoing evaluation of technologies to detect and track small unmanned aircraft systems (sUAS) in dense urban environments for military, and potentially homeland security, applications is making progress as the program enters its next phase, a Defense Advanced Research Projects Agency (DARPA) official tells HSR.

The second phase of Aerial Dragnet concluded in October with five days of testing in Southern California and DARPA and its industry partners are still analyzing the data but so far, the program has shown progress in being able to distinguish small drones from other moving objects, says Dr. Paul Zablocky, the agency’s program manager in an interview in late November.

“The jury is still out in terms looking at all the data but we’ve made a lot of progress on the classification algorithms to separate out various moving objects and we’ll continue to refine, but like I said we made quite a bit of progress there,” he says.

Aerial Dragnet began in 2016 with analysis of the challenges of detecting sUAS in dense urban environments and development of sensors to meet these challenges. The initial phase also included testing on military test ranges and other locations but these didn’t replicate dense urban environments, Zablocky says.

The second phase involved refining the algorithms to be able to detect drones “and not everything else that might show up in a very noisy environment” such as birds, he says. The evaluations in October in San Diego and nearby National City were the first in a dense urban environment. In addition to coping with the challenges of discriminating drones from other moving objects, there are also challenges in tracking sUAS as they move between and behind buildings.

The testing in Southern California included acoustic, optical and radar sensors. DARPA’s partners included Echodyne, Lockheed Martin’s [LMT] Advanced Technology Laboratory, Teledyne Technologies [TDY], and the Univ. of Washington’s Applied Physics Laboratory (APL-UW). Sensors were installed on two large tethered aerostats about 500-feet above the ground over the two cities, radar and optical sensors aboard hovering drones and at fixed building-top locations for wide area surveillance.

DARPA used about a dozen commercial-off-the-shelf sUAS to simulate unauthorized and unidentified drones flying below 400 feet in the urban areas

In addition to demonstrating improved detection capabilities, Zablocky said executing the evaluation in an urban environment was a challenge in itself. He said other participating organizations, which included the Department of Homeland Security, Federal Aviation Administration, various local law enforcement agencies, a special Army group, and the Navy in San Diego, were impressed the execution at scale in an urban environment and wanted to know how it was pulled together and implemented.

Echodyne Participation

More than a dozen of Echodyne’s radars were installed as part of the large area, comprehensive surveillance of drones under Aerial Dragnet.

Echodyne employed its EchoGuard and EchoFlight radars for the evaluation, which the company touted as successful.

“The DARPA requirement to create full urban airspace situational awareness has been challenging yet rewarding,” Tom Driscoll, Echodyne’s chief technology officer, said in a statement in November. “In conjunction with APL-UW, we operated more than a dozen radars on aerostats and rooftops to detect and track urban drone flights. Our performance demonstrated that Echodyne’s innovative, beam-steering, electronically scanning radars have unique operational, sensitivity, and intelligence characteristics necessary to conduct networked airspace surveillance over a major U.S. city like San Diego.”

Echodyne said the testing assessed how well the system could detect, track and identify over 150 sorties of drones.

As DARPA and its partners continue to process the data collected from the recent evaluations, Aerial Dragnet has entered the 15-month third phase which will include several more multi-day data collections in dense urban environments to “really make sure these algorithms are robust in a variety of applications,” Zablocky says. Returning to San Diego is an option, he says.

Phase 3 will also allow DARPA and its partners to better understand the performance of the systems and the performance of the “aggregate set of sensors as well,” Zablocky says.

Zablocky says that the sensors being used in Aerial Dragnet could probably be successful in conducing wide area surveillance of drones in open areas, including around airports in general. He highlights that the testing in San Diego included nearby San Diego International Airport. He says there is “good work” already being done for detecting and tracking drones in open areas or looking up into the sky.

“Hopefully other environments are less challenging for the set of sensors we’re looking at,” he says.