July 27, 2005 By Sherry Watkins
A car approaches, but the driver is unaware of the obstruction -- a split second before impact, he catches a flash of shaggy brown fur in his headlights. Brakes squeal as the car careens into the animal's legs, sending its bulk over the hood and smashing through the windshield.
There are no survivors.
These collisions have several states asking how to prevent such tragedies, and concern about them helped launch a study to explore options for alerting drivers that a large animal could be blocking the roadway ahead.
The study, Animal Vehicle Crash Mitigation Using Advanced Technology, is being conducted by the Western Transportation Institute (WTI) at Montana State University, Bozeman, but is funded by 15 different transportation departments and the Federal Highway Administration.
Each contributor hopes to gain insight into the feasibility of installing animal detection systems along roadways to reduce vehicle collisions caused by animals blocking the right of way.
This isn't the first study using animal detection systems -- since 1993, Switzerland has installed several.
"Data from the seven sites in Switzerland showed an average reduction in ungulate-vehicle collisions of 82 percent," said Marcel Huijser, research ecologist at WTI and lead on the project. "These results are quite encouraging, but we still have some research questions with regard to the factors that may make a system more or less effective.
"For example, drivers are more likely to lower their speed if road and weather conditions are poor, and the warning signs may have to be accompanied by advisory or mandatory speed limit reductions," he continued. "The goal of the study is to evaluate the reliability and effectiveness of these animal detection systems, regardless of the outcome, positive or negative."
Crossing the Line
Animal detection systems use two primary technology types: area-cover sensors and break-the-beam sensors.
Area-cover sensors register the presence of large animals within a certain range of the sensor using infrared light or microwave radio signals. The sensors can be enhanced with specific algorithms capable of distinguishing between large animals and other moving objects that could potentially generate false detections, such as vehicles or hot pockets of air, Huijser explained.
Break-the-beam sensors are triggered once an animal breaks the beam sent between a transmitter and receiver, and includes infrared, laser or microwave radio signals.
The WTI system in place in Montana is located on a stretch of Highway 191 at Yellowstone National Park. This break-the-beam system uses low-power signals -- approximately 35.5 GHz. When an animal's presence breaks the beam, the signals transmitted to the receiver decrease, triggering the system and setting off warning lights.
This system requires direct line of sight, and the transmitter and receiver must be within one-quarter mile of each other. Solar panels provide the power for each, and excess power from the panels is stored in batteries, which lend power in the dark.
Upon animal detection, an ultra high frequency (UHF) signal is sent from the main station to the four stations nearby, causing the warning lights to flash.
"Within this one location at Yellowstone National Park, we've got a lot of elk movement, and it's clear they cross most in one section of the system," Huijser said. "They seem to prefer certain areas more than others. But this preference has not been very absolute, and where animals cross [the road] also depends on the species, because different species have different habits."
The system monitors detections, sending them via UHF radio signal in real time to the master station to be saved. This way, all activity can be monitored with a portable data radio at the actual system site.
The Montana system is customized to accommodate physical traits of the location, as well as habits of the animals living there. With different systems available, it's possible to create unique alternatives for separate implementations.
"Different people have installed different systems for different reasons with different expectations," Huijser said.
Follow the Leader
Some animal detection systems use unique technology to accommodate unique situations. For instance, a system in Sequim, Wash., near the Olympic Peninsula uses radio collars on elk to trigger warning lights along the road.
Huijser said an elk herder identified the leaders of the herd, who were then outfitted with radio collars under the assumption that where they lead, the others will follow.
"This is a unique situation, where you have a resident elk herd present year-round, and you are also able to identify leading individuals in the herd," Huijser explained. "You also have a species that is highly gregarious, as it operates in large groups."
Four receivers with directional antennas spaced 1mile to 1.5 miles apart along Highway 101 continually scan for the frequencies set on the radio collars. Once a receiver detects an elk signal, flashing warning lights are activated to alert oncoming drivers. "The directional antennas can pick up signals ahead within a 60 degree cone up to the next antenna," explained David Rubin, a retired physicist and member of the Sequim Elk Habitat Committee who created the system as a volunteer.
To prevent false detection, the receivers are equipped with a device designed by Rubin that weeds out interfering signals generated by other sources. "By counting the radio pulses having appropriate pulse lengths, only elk transmitters trigger the lights," he said. Currently nine elk are outfitted with radio collars in Sequim.
Before the system's implementation in Sequim, an average of 2.5 elk-vehicle collisions were reported annually between 1994 and 2000. In the four-year span following the installation, there has only been one such collision.
"The decrease in elk-vehicle collisions is also noteworthy since elk have been crossing Highway 101 more frequently in the past four years, the volume of traffic on Highway 101 increases yearly, and the Highway 101 bypass allows higher speeds for traffic," said Shelly Ament, wildlife biologist from the Washington Department of Fish and Wildlife, who was responsible for coordinating the system's development.
Alerting All Drivers
For the past three years, a system in western Maine that's triggered not by animals, but by oncoming vehicles that cross the beam sent between transmitter and receiver, has been in place. The system alerts every car that breaks the beam with yellow flashing lights coupled with a moose caution sign.
"The main reason we've gone with vehicle-activated signs as opposed to animal-activated signs is there is potential for the latter to provide a false sense of security," explained John Perry of the Maine Department of Transportation's environmental office.
He said if a large animal, such as a moose, were to break a beam sensor, the triggered warning would eventually reset, and if the moose remained in the roadway, approaching vehicles would no longer receive a warning and would assume the path was clear.
"Our thought is that with the vehicle activating the sign, the lights would always flash, providing they were functioning properly, thereby putting the driver on alert.
"Maine DOT has not ruled out the use of animal-activated systems," Perry said. "We are looking into the feasibility of joining the two different systems [both animal and vehicle-activated signs in the same high-collision corridor], so that if the above situation occurred, the vehicle-activated signs should, in theory, alert the driver to reduce speed in any event."
Animal-vehicle collisions, particularly collisions with the largest mammals, are an important issue in Maine, he continued, with the main species of concern being moose, deer, and to a lesser degree, black bears. Moose can grow to 10 feet tall and weigh 1,500 pounds, making them a formidable obstacle.
From 1999 to 2003, a substantial number of accidents involving large animals have been reported in Maine:
In addition to the human-life factor, the economic loss is also an important issue in Maine. Between 1999 and 2003, according to the DOT, the estimated economic loss resulting from moose collisions was more than $76 million; more than $69 million from deer; $616,000 from bears; and almost $10 million from all other animals.
Although Perry said it's too soon to tell whether the system is having a significant impact, research efforts will continue as new data is collected.
Taking a Test Drive
Currently the WTI uses driving laboratories or simulator environments to gauge how people react to large animals and different types of warning signals, said Huijser, because the best way to devise new systems is to use such labs to replicate the environment being targeted.
"They can actually drive through the landscape on the road, and you can program animals jumping out in front of them and warning signals that may or may not be activated," he said.
In doing this, system implementers can decide on customized solutions based upon landscape, visibility and certain road characteristics. There are numerous variables used to influence a system's configuration, so long-term research is key. As Huijser noted, there's no single answer, and determining the best solution for any given area takes time.
"By having more sites, we will get more and better data on system effectiveness," Huijser explained, noting that new animal detection systems are planned for the next one to two years in California, Arizona and Wyoming.
Meanwhile, states with systems currently in place hope to save some lives in the process.
You may use or reference this story with attribution and a link to