Long-combination vehicles (LCVs) have significant potential to increase economic productivity for shippers and carriers by decreasing the number of truck trips, thus reducing costs. However, size and weight regulations, triggered by safety concerns and, in some cases, infrastructure investment concerns, have prevented large-scale adoption of such vehicles. Information on actual crash performance is needed. To this end, this work uses standard and heteroskedastic ordered probit models, along with the United States' Large Truck Crash Causation Study, General Estimates System, and Vehicle Inventory and Use Survey data sets, to study the impact of vehicle, occupant, driver, and environmental characteristics on injury outcomes for those involved in crashes with heavy-duty trucks. Results suggest that the likelihood of fatalities and severe injury is estimated to rise with the number of trailers, but fall with the truck length and gross vehicle weight rating (GVWR). While findings suggest that fatality likelihood for two-trailer LCVs is higher than that of single-trailer non-LCVs and other trucks, controlling for exposure risk suggest that total crash costs of LCVs are lower (per vehicle-mile traveled) than those of other trucks.

Abstract not found

DTD APPLIED RESEARCH AND INNOVATION BRANCHThe contents of this report reflect the views of the author(s), who is(are) responsible for the facts and accuracy of the data presented herein. The contents do not necessarily reflect the official views of the

Sensys Networks ’ VDS240 vehicle detection system is a wireless sensor network composed of a collection of 3 ” by 3 ” by 2 ” sensor nodes put in the center of a lane and a 6 ” by 4 ” by 4 ” access point (AP) box placed 15 ’ high on the side of the road. A node measures changes in the earth’s magnetic field induced by a vehicle, processes the measurements to detect the vehicle, and transfers the processed data via radio to the AP. The AP combines data from the nodes into information for the local controller or the Traffic Management Center (TMC). An AP communicates via radio directly with up to 96 nodes within a range of 150’; a Repeater extends the range to 1000’. This range makes it suitable to deploy VDS240 networks for traffic counts, stop-bar and advance detection, and measurement of queue lengths on ramps and at intersections, as well as parking guidance and enforcement. VDS240 is self-calibrating, IP-addressable and remotely monitored. Data are not lost because unacknowledged data packets are retransmitted. The accuracy of VDS240 for vehicle counts, speed and occupancy is comparable to that of welltuned loops. Because the nodes report individual vehicle events, the AP also calculates individual vehicle lengths, speeds and inter-vehicle headways—measurements that can be used for new traffic applications. In July 2007, VDS240 systems were deployed in arterials and freeways in several cities and states, and 30 customer trials were underway in the U.S., Australia, Europe and South Africa.

I would like to thank Dr. Christopher M. Monsere, for his advice and encouragement

8 Sensys Networks ’ VDS240 vehicle detection system is a wireless sensor network composed of a collection of 3 00 by 3 00 by 9 2 00 sensor nodes put in the center of a lane and a 6 00 by 4 00 by 4 00 access point (AP) box placed 15 0 high on the side of the 10 road. A node measures changes in the earth’s magnetic field induced by a vehicle, processes the measurements to detect the 11 vehicle, and transfers the processed data via radio to the AP. The AP combines data from the nodes into information for 12 the local controller or the Traffic Management Center (TMC). An AP communicates via radio directly with up to 96 nodes 13 within a range of 150 0; a Repeater extends the range to 1000 0. This range makes it suitable to deploy VDS240 networks for 14 traffic counts, stop-bar and advance detection, and measurement of queue lengths on ramps and at intersections, as well as 15 parking guidance and enforcement. VDS240 is self-calibrating, IP-addressable and remotely monitored. Data are not lost 16 because unacknowledged data packets are retransmitted. The accuracy of VDS240 for vehicle counts, speed and occu-17 pancy is comparable to that of well-tuned loops. Because the nodes report individual vehicle events, the AP also calculates 18 individual vehicle lengths, speeds and inter-vehicle headways—measurements that can be used for new traffic applications. 19 In July 2007, VDS240 systems were deployed in arterials and freeways in several cities and states, and 30 customer trials 20 were underway in the US, Australia, Europe and South Africa.

Long-combination vehicles (LCVs) have significant potential to increase economic productivity for shippers and carriers by decreasing the number of truck trips, thus reducing costs. However, size and weight regulations, triggered by safety concerns and, in some cases, infrastructure investment concerns, have prevented large-scale adoption of such vehicles. Information on actual crash performance is needed. To this end, this work uses standard and heteroskedastic ordered probit models, along with the United States ’ Large Truck Crash Causation Study, General Estimates System, and Vehicle Inventory and Use Survey data sets, to study the impact of vehicle, occupant, driver, and environmental characteristics on injury outcomes for those involved in crashes with heavy-duty trucks. Results suggest that the likelihood of fatalities andLemp, Kockelman, and Unnikrishnan 2 severe injury is estimated to rise with the number of trailers, but fall with the truck length and gross vehicle weight rating (GVWR). While findings suggest that fatality likelihood for twotrailer LCVs is higher than that of single-trailer non-LCVs and other trucks, controlling for exposure risk suggest that total crash costs of LCVs are lower (per vehicle-mile traveled) than those of other trucks.

Long-combination vehicles have significant potential to increase economic productivity by decreasing the number of truck trips and thus reducing costs. However, size and weight regulations imposed due to safety concerns and, in some cases, infrastructure investment concerns have prevented large-scale adoption of such vehicles. Information on actual crash performance is needed. To this end, this work uses standard and heteroscedastic ordered probit models to study the impact of vehicle, occupant, driver and environmental characteristics on injury outcomes for those involved in crashes with heavy-duty trucks. Results suggest that the likelihood of fatalities and severe injury is estimated to rise with the number of trailers, but the presence of an LCV can more than compensate for this trend. In addition, injury severity was found to decrease when driving under well lit conditions, on freeways and in the absence of grades.

This report describes the development of real-time crash prediction models for the Interstate-4 corridor in Central Florida area. Crash data for 36.25-mile freeway stretch from the year 1999 through 2002 has been used to link the crash occurrences with real-time traffic patterns observed through loop detector data. The analysis technique adopted for this phase of the study is with-in stratum matched case-control logistic regression. The purpose of matched case-control analysis is to explore the effects of independent variables of interest on the binary outcome while controlling for other confounding variables through the design of the study. In the context of this research crash or non-crash is the binary outcome with traffic parameters being the independent variables. The design of the study allows controlling for external factors such as geometric design of the freeway, time of the day, day of the week, etc., and hence they are implicitly accounted for them. Using this technique two types of models, i.e., simple and multivariate, were developed. Prior to development of the models some of the data related issues such as data cleaning, determination of exact time of the historical crashes, etc. were addressed. Both types of models were evaluated based on their classification performance. It was observed that although the simple models have