On average, nine collisions occur each day in the U.S. between trains and road vehicles. These collisions result in fatalities forty times more often than for other vehicle accidents. Eighty percent of collisions occur at railroad crossings that do not have train-activated warning systems. Using conventional equipment to upgrade these 80,000 public crossings will cost over $10 billion, and take more than eighty years at the present rate. In addition, there are 95,000 private railroad crossings that lack train-activated warnings. This research will develop a low-cost, train-activated warning system incorporating: a novel, proprietary train sensor; alternative warning technologies; common, lower-cost traffic signal components; secure wireless networking between the sensor and signal lights to reduce installation costs; and solar power (with substantial battery reserves) to service off-grid locations. A major objective is optimizing visual and audible cues to provide a safety benefit competitive with crossing gates while avoiding the cost and power requirements of gates. Additional warning cues to be studied include in-pavement lights, variable message signs, active advance warning signs prior to the crossing, and electronically synthesized bells and train horns. The proposed system would greatly accelerate the rate of elimination of dangerous crossings that are predominantly in rural areas. The estimated train arrival time broadcast by the train sensor could also be used to generate in-vehicle alerts for school buses, HAZMAT trucks and emergency vehicles. OBJECTIVES: The aim of this research is to develop an alternative, low-cost, train-activated warning system to improve safety at rural railroad crossings. Conventional train-activated crossing warning systems use gates and flashing lights. These systems are effective, but the cost per crossing severely limits the rate at which unprotected crossings can be upgraded, and substantial power requirements necessitate access to the power grid. The proposed warning system will use additional audiovisual cues, alternative construction techniques, and wireless networking between the sensor and warning devices to achieve competitive levels of safety improvement, but with reduced costs and power requirements. For sites without easy access to the power grid, the proposed low-power approach should further reduce costs by enabling the use of solar power (with appropriate battery reserves). The basis of this proposal is that diverting some funding from existing budgets for crossing upgrades to the proposed low-cost systems could provide a greater net safety improvement by promoting a more rapid elimination of unprotected, high-risk public and private crossings. To establish feasibility in Phase I, the first objective is to qualify a set of alternative warning devices and other system components, and then estimate the cost-benefit ratio of using various combinations of these devices, relative to the conventional approach. Within this objective, each aspect of the conventional warning system will be analyzed along with the possible alternatives. This will include the train sensor; methods of visual and audible warning at the crossing; the possible benefit of advance active-warning signs prior to the crossing; the traditional heavy-duty construction methods, materials and components employed in the railroad industry compared to the more common and cost-effective engineering employed for traffic lights; applicable standards and regulatory requirements; maintenance methods and costs; and overall system power requirements. The second Phase I objective is to specify and cost a secure wireless data link between the train sensor and the various warning devices based on a commercial standard such as 802.11. The aim is to reduce installation costs associated with trenching and cabling, and reduce maintenance costs by allowing a passing train or road maintenance vehicle to test and monitor system performance. The third Phase I objective is to prepare detailed physical descriptions and behavioral modeling for each alternative system. This information will be used in Phase II to develop simulator models for the University of Minnesota's vehicle simulator, so that the most promising systems can be tested and optimized. Based on those results, the most cost-effective warning system will be selected for prototype construction. Extensive field trials will then be performed at the Transportation Technology Center, Inc., near Pueblo, CO, to verify component and system reliability. Successful demonstrations of the system to commercial partners and potential customers, such as federal and state DOT representatives, will then support Phase III commercialization objectives. APPROACH: In Phase I, the Project Director (PD) will analyze prices and equipment specifications for commercially available equipment that would be suitable for the proposed warning system. Warning devices, sign components and strategies to be investigated include: low-power LED replacement bulbs, low-cost traffic light poles and mounting hardware; variable text message signs, in-pavement flashing lights; and strobe lights or train-activated lights on the advance crossing warning sign. Additional audible cues will also be considered such as electronically simulated crossing bells and train horns, with options for highly directional speakers in populated areas. Since 25% of collisions involve the road vehicle colliding with the side of the train, methods of illuminating the train such as flood lights or laser pattern generators mounted on the back of the warning sign will also be investigated. Changing messages or light patterns will be considered to provide an indication of the train presence or direction and time to arrival. As part of the aim to further reduce installation costs in remote locations by providing for off-grid operation, the power budget of the selected components will be estimated for various rail traffic densities. The generation and storage characteristics of commercially available solar panels and batteries will be analyzed in consideration of US regional requirements. Candidate warning systems will be designed based on the most promising combinations of technologies, and total system costs will be estimated. Consultant Dr. T.J. Smith of the University of Minnesota's School of Kinesiology, Human Factors Research Laboratory, will estimate the efficacy of each candidate warning system based on a review of existing traffic research and statistics, and his previous research on railroad crossings. Estimated accident reduction rates will be used to calculate a cost-benefit ratio for each alternative system which will then be compared to the conventional approach to decide feasibility. These designs and specifications will be used in Phase II to develop simulator models to test, optimize and validate the estimations above using the University's vehicle simulator. The most cost-effective warning system will be selected and a prototype will be developed for Phase II field trials. Wireless networking between the train sensor and warning devices will be investigated, based on commercial WiFi networking standards. The range and communications protocol will be designed to allow drive-by testing of the crossing system. Possible use of the train sensor transmission to enable in-vehicle warnings for school buses, HAZMAT trucks and emergency vehicles will be considered. Wireless networking standards, costs and security will be addressed. Railroad and highway regulations will be studied to ensure compliance and to investigate the possibility of placing the active warning signs off the railroad right-of-way. A proprietary train detection method based on Time Domain Reflectometry, developed under separate research funding through the National Academies, Transportation Research Board, will be characterized for use in the present application
