Consider dynamic natural conditions and surrounding landscapes when selecting technologies for animal detection system deployments.

Montana and Pennsylvania’s experience with deploying animal detection systems.

Date Posted
11/19/2007
TwitterLinkedInFacebook
Identifier
2007-L00407

Animal Vehicle Crash Mitigation Using Advanced Technology Phase I: Review, Design and Implementation

Summary Information

Animal-vehicle collisions affect human safety, property and wildlife. The number of these types of collisions has increased substantially over the last few decades. This report describes the results of a project that explored the prospects for a relatively new mitigation measure to reduce animal-vehicle collisions: animal detection systems. Animal detection systems use ITS to detect large animals when they approach the road and to warn oncoming vehicles of their presence.

The report identifies existing animal detection system technologies and describes the selection of two experimental detection systems and their installation at two field sites:

  • US Highway 191, in Yellowstone National Park, Montana. This site stood out because of its national visibility, representative terrain and vegetation (forested hills and mountains), the abundance of large mammals, and its proximity to the office of the Western Transportation Institute at Montana State University.
  • Highway 22/322, near Thompsontown (northwest of Harrisburg) Pennsylvania (PA). This site ranked high because of the large number of animal-vehicle collisions; controlled access; limited fluctuations of deer populations; relative proximity to the vendor; and the terrain, vegetation and large animals seemed representative for eastern states.

This report documents experiences and lessons learned with planning and design, installation, operation and maintenance, test results on reliability, and user acceptance of animal detection systems.

Lessons Learned

By providing drivers with advanced warning, animal detection systems reduce the likelihood of animal-vehicle collisions on roads in or adjacent to wildlife habitat. As the majority of large mammals are found in undeveloped lands, away from urban centers, animal detection systems are often located in areas where the natural environment poses unique operational challenges. In order to alleviate the effects of dynamic natural conditions, the following considerations and actions need to be taken prior to, during, and after deployment of animal detection systems.

  • Employ solar power as a cost effective means for powering animal detection systems. Most animal detection systems are located in rural and remote areas where 110 V or 220 V, if available, may be relatively expensive to bring to individual sensor stations.
    • The cost for trenching and conduit alone, to two sides of the road, over a half mile, was estimated to be a minimum of $52,983 at the site near Thompsontown, PA. In contrast, the total cost for solar power was estimated at $7,500.
    When deploying Solar power technology agencies should position panels so that shadows are taken into account, especially during winter when the sun is low on the horizon and shadows last longer. Battery storage should also be designed to cope with long nights, dark days, and potential snow cover. For areas that do not receive enough sunlight, such as deep canyons or the arctic, fuel cell technology may be an option.
  • Anticipate the occurrence of false positive and false negative detections and plan accordingly. False positives occur if the system is triggered by causes other than the presence of large animals and false negatives occur if a large animal is present, but the system fails to detect it. Frequent false positives may cause drivers to eventually ignore activated signs, while frequent false negatives present drivers with a hazardous situation. False positives and false negatives may have a wide variety of causes, though the systems at the Montana and Pennsylvania sites suffered especially from the causes shown below:
    • False positives
      • Faulty hardware
      • Hardware that does not meet the requirements (e.g., temperature range)
      • Incorrect signal strength
      • Inability of the software to distinguish between true and false detections
      • Software errors
      • High, moving and wet vegetation
      • Passing traffic
      • Heavy snowfall with high moisture content
      • Snow spray from snowplows and other vehicles
      • Snow and ice accumulation in front of the sensors
      • Snow accumulating on the ground and covering the sensors;
      • Vehicles turning on and off the road
      • Vehicles parked in the right-of-way; and
      • Accumulating snow on the ground, blocking or bouncing signals.
      False negatives
      • Incorrect location or height of sensors, leading to blind spots in curves or low areas in the right-of-way;
      • Faulty hardware
      • Incorrect software definitions (settings) for valid detections;
      • Snow accumulation on ground leading to incorrect sensor height at high snow levels
      • Sensors that are out of alignment (possibly due to expansion and shrinking of wooden poles with varying temperature and moisture contents
      • Vandalism or perhaps the rubbing against equipment by large animals, (e.g., elk or bears).
    Post installation testing at the Montana site revealed numerous false positives. In an effort to reduce the number of false positives, software filters were installed to distinguish between detections caused by vegetation, passing vehicles, and snow versus large animals. However, snow spray from snowplows can still trigger the system.
  • Provide for remote access to optimize the deployment of resources. While not essential to the functioning of an animal detection system, remote access to gather and analyze data and other parameters (e.g., battery voltage, output of solar panels) will save time and money as it allows basic system functions to be checked without conducting site visits. Remote access can allow for the disconnection of certain sensors or stations if an isolated problem, that cannot be immediately resolved, is detected. The functioning sensors will then remain in operation and will warn drivers when large animals approach the road in the sections with the active sensors or stations. In addition, remote access may allow for the connection or disconnection of the flashing warning signs and changing the time that the lights flash after the last detection has occurred. Data logging can be accomplished by allowing the animal detection equipment to periodically and autonomously download to an online database. Under this procedure, the detection equipment could produce reports whenever the data link is available. These data services are less expensive than “connection on demand” services and are readily provided by low earth orbiting satellites (LEOS) and cell phone systems.
    • The system at the Montana site stored detection data on a flash card including: radio link failures, radio link signal levels, beam break summaries, logging memory status, station identification number and sensor, whether the flashing lights were turned on or not, battery voltage, solar panel voltage and temperature in the box with equipment. The last three parameters allowed for identifying potential problems. In addition, the logging system maintains and reports statistics associated with the operation of individual elements of the system. All data can be downloaded on-site or from a remote location through a modem and land-based phone line. The time delay for signal turn off could also be set from a remote location.
    • In Pinedale, Wyoming a camera and recording system are used to help validate and calibrate the animal detection system. The camera system allows for remote monitoring from the local WYDOT office.
    • At the animal detection site in Nugget Canyon, WY three separate systems are used which are all linked to one of three traffic counters to allow for remote access to the data.
  • Ensure adequate coverage can be achieved when planning for remote access. Remote access can be obtained through a modem and a telephone connection (land based, cellular or satellite) or radio signals if one is in the immediate area. However, agencies deploying communications technologies need to take into account certain limitations and plan accordingly. For example, if connecting via cellar telephone thorough testing for signal strength is required, as coverage can be poor or variable, especially in remote and rural areas. Establishing what technologies are available based on any limitations, will allow agencies to deploy a communications system that will provide full functionality from the onset of its installation.
    • At the Montana site cell phone coverage was insufficient for reliable data transmission. The poor performance of the telephone line connection to the Roadway Animal Detection System (RADS) encouraged the exploration of alternate communication methods for remote access. As a result a satellite based system was found which was tailored to this application. At the time the report was written it was the intention that this system would be demonstrated at the site along US Hwy 191 in Yellowstone National Park.
  • Prevent obstruction of detection sensors by mitigating the effects of snow accumulation. The accumulation of snowfall on the ground and on RADS sensors can lead to a number of challenges in detecting the presence of large animals. Another factor that can cause problems for sensors is snow spray from snowplows and other vehicles. When deploying animal detection systems in areas where snowfall is possible, agencies need to seek methods or tools to ensure sensor detection will go unhindered during snow events.
    • At the Montana site, because accumulation of wet snow can reduce the signal level of beams, radomes (shields) were added to the transmitter and receiver stations to prevent the accumulation of snow and ice in the beam tube aperture.

In many cases the need for animal detection systems exists in areas relatively cutoff from large urban centers. As a result common technologies and resources may not be readily available. Additionally, it is not uncommon for sites to be located in rural and remote areas, where climate conditions and surrounding natural landscapes pose challenges to the successful deployment and operation of animal detection systems. These challenges vary and can be overlooked in the absence of careful planning and evaluation. Agencies deploying animal detection systems need to plan and adjust for natural conditions that affect power supply, communications, data access, system reliability, and sensor detection. Following this guidance in implementing animal detection systems will help to support the ITS goal of safety.