CEE researcher works to reduce bird-plane collisions

By Laura Weisskopf Bleill

For much of the past decade, Professor Edwin Herricks has laid the foundation for a tool that may address one of the oldest problems facing aviation today, birds colliding with aircraft. 

Herricks, Professor of Ecological Engineering, coordinates the Airport Safety Management program for the CEE-based Center of Excellence for Airport Technology.  His team has deployed avian radars at major airports around the country.  Now Herricks’ research program is moving into the performance assessment phase, addressing issues such as calibration, performance under different environmental conditions, and reliability. A radar at the Naval Air Station at Whidbey Island in Washington State has been operational and collecting data for more than 18 months. A radar at Seattle-Tacoma International Airport and a second one at Whidbey Island have been providing data since August 2007.

The ultimate goal is to provide guidance to the Federal Aviation Administration (FAA) so they can develop procedures that may help pilots, air traffic controllers and wildlife managers avoid a catastrophic bird strike accident.

Already, bird strikes cost the aviation industry $1 billion a year, according to Bird Strike Committee USA. But there are few mandates in place to address the issue.

“In one sense the FAA is trying to get out in front of the avian radar question,” Herricks said. “Before the FAA requires the installation of this technology, they need to figure out what it can do.”

At Whidbey Island, one radar uses a parabolic dish antenna, while the other has an array antenna. At Seattle-Tacoma, one installation has two dish antenna radars sitting on top of an administration building, and a radar with an array antenna is used on the airfield.  The difference between array and dish antennas is beam coverage.  The array antenna sends out a signal that sweeps plus or minus 11 degrees from horizontal, providing coverage from the ground to 11 degrees up from horizontal.  The array configuration provides information on the range to target, but no altitude information.  The parabolic dish antenna sends out a four-degree cone with an angle adjustment that provides better altitude discrimination.  The radars generate data that can show the real-time movements of birds around the Seattle Airport.  A visual display of bird movements is available on Google Earth. 

“These radars with different antenna configurations and different ways that we’re using them represent different classes of operation and information that we need for this overall performance assessment,” Herricks said.  “We already know that operations are significantly different between Seattle and Whidbey Island simply because of the differences in the numbers of birds at Whidbey Island. So what we need to do is evaluate these sensors under a variety of conditions to see how they perform under all of these conditions.”

One of the challenges is to determine that the tracks the radar picks up are indeed birds. The radar data processor can be configured to use factors like velocity to discriminate between birds and other aerial targets, such as insects and rain. 

The sensor validation work is ongoing. Herricks recently flew a radio-controlled helicopter at Seattle-Tacoma, an experiment that confirmed that the radar can detect a known target in near-real time.

Another challenge is sifting through the sheer amount of data that has been collected, which Herricks compares to “drinking water out of a fire hose.” In just a year the operational radars have produced a terabyte, or 1,024 gigabytes, of processed data.  But Herricks believes it is possible to provide information on wildlife hazards that will be useful to pilots.  One approach is to automate the analysis of tracks of birds flying across approach or departure zones at airports.     

“By counting those bird tracks, what we begin to do is develop a quantitative measure of hazard and we can do this throughout the day so we can begin to identify critical periods, taking into account  changes that occur between days and changes that occur between seasons. This will help pilots and controllers understand how the relative risk of collisions with wildlife changes over time.”

As part of the FAA program Herricks’ team will be deploying avian radars at Chicago’s O’Hare International Airport, New York’s John F. Kennedy Airport, Vancouver International Airport, and Dallas/Fort Worth Airport.  Herricks estimates that the evaluation process may take another two to three years, to assemble the data needed for the FAA to develop an advisory circular on avian radars.

In addition to the avian radar project, Herricks is spearheading a process to assess foreign object debris (FOD) detection systems at airports across the nation. The goal is similar—to provide the FAA with technical information so that it can develop an advisory circular that establishes the requirements and standards for the technology. 

Vendors worldwide have developed continuous surveillance systems that can detect FOD—the junk that collects on or around runways—which is said to cost airlines between $4 billion and $10 billion a year.  The FAA brought in Herricks’ team to assess the performance of four such systems. Testing has been completed for a stationary millimeter wave radar system at TF Green International Airport in Providence, R.I.  Testing is underway for a hybrid radar and video system at Boston’s Logan International Airport, as well as a mobile millimeter wave radar system and a high resolution video camera system at the major Chicago airports.

Although the sensor technology for each system is different, Herricks’ team uses common procedures and FOD targets. The performance assessment has four elements—calibration, sensitivity testing, blind testing and operational performance analysis. Testing on runways is a challenge.  Herricks’ team is commonly placing FOD items on runways in the middle of the night when runways are closed.  The assessment programs are designed to test detection systems under different weather conditions, with a particular emphasis on rain, ice and snow.

“The FAA needs to have information about these detection capabilities and a variety of other features of the technology so they can write this advisory circular,” Herricks said. “We’re carefully trying to avoid endorsement. That’s why we’re calling it a performance assessment, not an evaluation. We don’t want to evaluate. What we’re trying to do is simply use what the vendors are telling us is their detection ability and apply science to the test procedures to determine if detection claims are met.”

A tragedy accelerated the gears in motion for more sophisticated and specialized FOD detection systems, after investigators blamed a piece of debris for the 2000 Concorde crash in Paris. A titanium strip fell from a DC-10 that had departed from the same runway minutes earlier. This strip punctured one of the supersonic passenger jet’s tires during takeoff. The accident killed all 109 aboard the aircraft, in addition to four people on the ground. 

Herricks has reason to believe that the aviation industry is one commercial airline disaster away from making bird strike prevention a higher priority.  A bird strike made national news in March 2007, when a 767 hit a flock of canvasback ducks upon takeoff at O’Hare, ingesting several into one engine and shutting it down. The plane was able to return to the tarmac safely, but not without significant structural damage.

“We almost had the smoking hole in the ground that would have been very clearly a bird strike problem,” he said.

Herricks contends that the avian radar systems will significantly improve wildlife management at airports. The data will be able to identify trends that indicate when and where bird activity may increase, allowing wildlife managers to provide more concrete hazard warnings.

“At present what we have is the same warning day after day after day, and pilots ignore it,” he said. “We can actually begin putting information up there that is timely and real.”