6. HIGHWAY SAFETY APPLICATIONS

            This Chapter focuses on highway-safety discussions primarily on highway-rail grade crossings, pedestrian and bicycle safety, intersection safety, speed management, and work zone safety. As such, the opportunities for ADUS highway safety applications discussion in this report are those that can contribute to addressing issues facing the aforementioned safety areas. The opportunities for ADUS public safety applications and transit safety and securing application are discussed in chapters 4 and 7, respectively.

            Highway safety is perhaps the area that is targeted the least in terms of data archiving. Nonetheless, because highway safety is a complex interaction among vehicles, roadways, environments and drivers, highway safety is also the area that will benefit the most from data archiving. Since highway safety transcends geographic boundaries (metropolitan vs. rural), data archived from almost any of the ITS infrastructure components (e.g., traffic surveillance) can be used to meet some of the data requirements in highway safety applications. For example, volume data collected from an ITS traffic surveillance project can be used to estimate accident exposure by time of day and day of the week. Another example is archived speed data which, in conjunction with information on highway geometry and weather, can be used to estimate the propensity of incidents or accidents. That said, data integration becomes ever more imperative in highway safety applications.

6.1 STATE-OF-THE-PRACTICE REVIEW

            Almost all of the data elements collected by the Freeway and Arterial Management Surveys can be used to fill data gaps in highway safety, particularly the exposure aspects of highway safety. From the highway safety perspective, exposure information categorized by different vehicle types is equally important to information on overall vehicle exposure.

            Based on the information collected from the ITS Deployment Tracking Surveys, by far the most commonly collected freeway data are: traffic volume, and information on scheduled work zones (Figure 6.1). The most common technique used to collect traffic data is loop detectors, followed by video imaging detectors. Figure 6.2 depicts the prominence of different techniques used to collect traffic data. It is obvious from this figure that less intrusive technologies are becoming popular in collecting traffic data. Traffic volume and vehicle classification data which are essential in deriving vehicle-type specific exposure data are the two data elements that are most likely to be archived. Eighty-seven percent of the agencies in 1999 that collected traffic volume data also archive them. And, seventy-six percent of the agencies that collect vehicle classification data also archive them (Table 6.1).

Figure 6.1 Freeway Data Generation and Archiving

(Out of 74 Agencies Reported Data Generation and Archiving in 1999. 

Out of 66 Agencies Reported Data Generation and Archiving in 2000)

Figure 6.2 Techniques to Generate Traffic Data

1999 and 2000 ITS Deployment Tracking Surveys 

            * These questions were not asked in 2000

           Other data elements specifically pertinent to highway safety are:

            ●          Information on current and scheduled work zones. Almost three quarters of the agencies that collected this information archive it.

            ●          Weather and roadway conditions. Weather conditions and road conditions are archived by at least half of the agencies that collect/report weather and road conditions as part of their Freeway Management Systems as well as their Arterial Management Systems.

            ●          Information on ramp metering strategies help understand the propensity of side swipes and rear-end collisions on ramps.

            ●          Data archived on emergency responses and emergency vehicle signal preemption provide a foundation to evaluate the effectiveness of emergency management controls in saving lives.

            ●          Information on traffic signal controls (turning movements, phasing and cycle lengths and queues) advances knowledge about the causes of intersection accidents.

            ●          Pedestrian and bicycle traffic at intersections.

            Tables 6.2 and 6.3 show the propensity for data archiving in 1999 and 2000, with a significant downward trend. No agencies collect all seventeen data elements identified on the survey questionnaire. In 1999 one agency collected 14 out of the 17 data elements and it archived every data element collected. Overall, 31 of the 78 agencies in 1999 archived all of the traffic data collected (cells on the diagonal line) while the corresponding numbers are 19 out of 66 agencies in year 2000. Ten agencies in 1999 and eight agencies in 2000 reportedly collected traffic data but did not archive any (the “0“ column).

Table 6.2 Distribution of Agencies by Number of Data Elements Generated and Number of Data Elements Archived

1999 ITS Deployment Tracking Survey

Table 6.3 Distribution of Agencies* by Number of Data Elements Generated and Number of Data Elements Archived

2000 ITS Deployment Tracking Survey

* Excludes agencies that archived more data elements than they actually collected.

Usage and Users of the Archived Freeway Data

            Based on the tracking survey results, the media is the largest archived data requester/user, followed by state Departments of Transportation (Figure 6.3). The archived traffic data are primarily used for traffic analysis and planning (Figure 6.4). Sixteen agencies reported that archived arterial data were used for accident prediction models. Unfortunately, the specific applications of the analysis are unclear from the surveys. It could not be determined which of the users use which of the archived data and for what purpose. An enhancement to this survey would be to add “safety analysis” as a “primary use” of the archived data.

 Figure 6.3 Number of Agencies Responded to Freeway Data Requests by Requesting Institute

Figure 6.4 Arcvhived Freeway Data Usage

1999 and 2000 ITS Deployment Racking Surveys

            Three hundred eighty-eight agencies responded to the 1999 Arterial Management Survey. One hundred eighty-one of these agencies reportedly collected arterial data. Of those that reported data collection activities, more than two-thirds collected data on traffic signal controls (Figure 6.5). The percentage of agencies that archived their traffic signal data is very high (Figure 6.5).

Figure 6.5 Data on Traffic Signal and on Emergency Management

(from 388 Agencies Responding to the 1999 Arterial Management Survey)

          In addition to data collected/archived in Freeway and Arterial Management Systems, data that could be archived from Commercial Vehicle Operations and rural ITS infrastructure can support analyses of motor carrier safety and of highway railroad crossing accidents. For example, motor carrier safety can be enhanced through more efficient and timely information exchange. In-vehicle devices (such as those equipped in the ITS/CVO Technology Truck, see http://www.ornl.gov/dp111/index.htm) can minimize truck rollover and collisions, and an in-vehicle signing system can alert school bus drivers of potentially dangerous railroad crossing situations.^6.1 Data archived from these in-vehicle systems can improve our knowledge of pre-crash events, thereby improving the development of preventive counter-measures.

            Technologies and control strategies are available that can coordinate traffic signals and notify vehicles of approaching trains at intersections. There are a few pilot projects: (1) Vehicle Proximity Alert System at Colorado, and (2) Smart Railroad Highway Crossing project at Long Island, New York. The Vehicle Proximity Alert System used an in-vehicle warning (audio/visual) to warn drivers of priority vehicles (e.g., emergency vehicles, school buses, and hazardous material haulers) of approaching trains at highway-rail grade crossings. The second project, the Smart Railroad Highway Crossing project at Long Island, New York, is to develop and test a prototype integrated warning system. None of these projects considered the idea of archiving operational data.

            A limited amount of information on highway-rail intersections was asked in the Arterial Management survey. Three hundred and eighty-eight agencies were asked whether they deployed or plan to deploy by 2005 any technology at highway-rail intersections. Three-quarters of them reported “yes.” However, only 10% of these agencies actually had safety-improving technologies deployed, while the remaining 90% are planning to do so by the year 2005. Among the existing deployments, the ability to predict train arrival times electronically is the greatest requirement (Figure 6.6).

Figure 6.6 Capabilities/Technologies Deployed at Highway Rail Intersections in 1999

(Based on Arterial Management Survey)

Almost all of those who responded indicated that they convey information on highway-rail crossings to travelers via roadside media. Ninety-five percent of those who responded receive blockage information at highway-rail intersections to manage incident response. No questions were asked in the Arterial Management Survey regarding whether information on highway-rail intersections is collected or archived. Our impression at this point is that no data were archived.

6.2 CASE STUDIES OF HIGHWAY SAFETY APPLICATIONS

            Highway safety involves complex interactions among vehicles, roadways, environments, and drivers. As such, highway safety is probably the area that will benefit most from data archiving. For example, almost all of the data elements collected by traffic surveillance ITS deployments (e.g., traffic volume, speeds, and traffic mix) are pertinent to highway safety, particularly in measuring exposure to highway crashes/incidents. The data provide both spatial and temporal information (the latter by time of day and day of the week). When integrated with archived weather and roadway condition data, hourly traffic volumes and crash data were used to develop a model that predicts the mobility and safety impacts of winter storm events in a freeway environment.^6.2

            Another reason why highway safety will benefit considerably from data archiving is that it is not limited by geographic boundaries (metropolitan vs. rural). Data archived from rural ITS deployments can be used to meet some of the data requirements in highway safety applications.

            Because highway safety is so complex, data requirements for highway safety are much more demanding and complicated than for other applications (i.e., transit, CVO, roadway, and rural). Consequently, the use of archived data to meeting highway safety information needs faces two challenges that are unique compared to those faced by other applications. First, since data on four different areas (i.e., drivers, vehicles, roadways, and environments) are needed, data integration is the key to successful ADUS safety applications. Second, data from these four areas are subject to different degrees of data confidentiality concerns. These concerns affect the nature and extent of data archiving and data sharing. For example, information on weather conditions has been archived and widely disseminated probably since the inception of the world wide web, while archiving and sharing information on drivers and vehicles demand great caution in terms of liability and privacy issues.

            Recognizing the importance of data integration and data sharing, transportation and public safety communities use electronic and wireless technologies to integrate and share information. Two of the leading efforts are highlighted in this report: the National Model for the Statewide Application of Data Collection & Management Technology to Improve Highway Safety project in Iowa, and the Capital Wireless Integrated Network (CapWIN) project in the Washington metropolitan area. A wealth of detailed information on innovative automated data collection technologies used by law enforcement and transportation communities can be found on the International Association of Chiefs of Police (IACP) website.^6.3

6.2.1    National Model for the Statewide Application of Data Collection & Management Technology to Improve Highway Safety Project

            The goal of the National Model is to demonstrate, in a statewide operational environment, how new technologies and techniques can be cost-effectively used to improve the collection and management processes of highway safety data. Specifically, the Model is to:

            ●    improve data acquisition related to roadway incidents,

            ●    use technology to assist law enforcement,

            ●    streamline the communication of safety information to key stakeholders, and

            ●    extend the use of this information by safety and law enforcement programs.

When this model is deployed nationwide, it is anticipated that the quality of the nation's safety data will be improved through software, communications, and technologies.

            The project partners believe that these technologies can reduce data collection time, thereby minimizing disruption to traffic and improving data quality. The partners also believe that this project will contribute to better informed highway safety decisions and improved safety of the highway system.

            The National Model is managed as a consortium effort. The project partners include the Motor Vehicle Division of the Iowa Department of Transportation, Iowa Department of Public Safety/Iowa State Patrol, FHWA Iowa Division, and sheriffs and police departments across the state.

            At the heart of this Model is the Traffic and Criminal Software (TraCS). TraCS consists of a mobile client Microsoft Windows® based application that allows law enforcement officers to collect, validate, print, and receive information in their vehicle using either a notebook or pen-based computer (Figure 6.7). Information gathered with TraCS Mobile can be transferred to the TraCS Office and TraCS Enterprise database applications for reporting, analysis and retrieval.  TraCS consists of five components. These are:

            ●    the Mobile Accident Reporting System (MARS),

            ●    the Vehicle Safety Inspection System (VSIS),

            ●    the Electronic Citation Component (ECCO),

            ●    the Mobile Operating While Intoxicated (MOWI), and

            ●    Complaint/Incident Report Form (CIRF).

Figure 6.7 TraCS Mobile Unit

Source: http://www.dot.state.ia.us/natmodel/indiex.htm

        Built around a pen-based computer, the MARS component of TraCS allows the officer to enter crash data at the scene using a combination of point-and-click lists and free form data. ECCO allows the officer to use a point-and-click list of violations and fines. Fines for the specific violation can be computed automatically, and the tickets are transmitted directly to the courts. Consequently, the courts will electronically send the conviction notices directly to the Iowa DOT.

            VSIS is the vehicle inspection form that is used to inspect commercial vehicles and is issued by Iowa DOT Motor Vehicle Enforcement Officers. The MOWI component allows for electronic transmission of the state’s OWI form to the DOT. An officer enters or captures an individual’s information which is shared with the other components of TraCS as needed.

            At the time of the writing of the report, this project uses 375 laptop computers, 225 pen-based computers, and numerous digital imaging bar code readers. The software, TraCS, is used in 100 local agencies statewide, and by all state patrols. During the last five years, the project has resulted in some outstanding improvements:

            ●    Reduction in the amount of time that incident information is electronically available to the state to as little as one day,

            ●    Elimination of duplicate data entry by law enforcement agency staff,

            ●    Significant reduction of errors on accident reports, and

            ●    Reduction in time spent completing accident reports.

            Vital incident information collected from TraCS include: demographic information on the persons involved, the attributes of the vehicles involved, the severity of the incident, road and weather conditions, and the location of the incident. These data are stored indefinitely in the TraCS enterprise database and are accessible by partnering agencies. The data have been used to make informed decisions, and to identify and address emerging safety trends such as aggressive driving. Specifically, archived TranCS have been used to:

            ●    Conduct before-and-after analyses to determine the safety improvement of construction projects,

            ●    Conduct analysis of railroad crossing crash data to determine if changes in traffic controls are necessary,

            ●    Determine the effectiveness and adequacy of traffic controls at intersections,

            ●    Develop and prioritize pavement resurfacing projects for safety enhancement,

            ●    Evaluate the safety effectiveness and performance of traffic control signing, channelization, and marking at work zones; and make appropriate changes immediately after problems arise. The Traffic Management Center in Des Moines is configured to use the archived TraCS data to ensure the safety and mobility during the 6-year reconstruction of I-235.

6.2.2    Capital Wireless Integrated Network (CapWIN) Project

            The public's concern for safety in the Washington Metropolitan area has generated a need for improved coordination and information sharing among public safety and transportation agencies in Maryland, Virginia, and the District of Columbia. To meet this need, public safety and transportation agencies are using technologies to share more timely and accurate information among agencies serving the Capital Beltway and surrounding Washington areas.

            CapWIN was created to build a bridge between transportation and public safety communities and is a partnership between the States of Maryland and Virginia, and the District of Columbia. This project is developing an integrated transportation and criminal justice information wireless network by integrating transportation and public safety data and voice communication systems. This project is the first multi-state transportation and public safety integrated wireless network in the United States. The CapWIN Project is sponsored by the following agencies:

            ●    Maryland State Highway Administration,

            ●    Virginia Department of Transportation,

            ●    U.S. Department of Transportation (FHWA),^6.4

            ●    National Institute of Justice, Office of Science and Technology, and                  

            ●    Public Safety Wireless Network .

            A pilot test will be initiated during the strategic planning phase of the project. The pilot will include a minimum of six, and up to fourteen, in-vehicle systems that will allow messaging between police vehicles in Maryland, Virginia and Washington, D.C.; transportation vehicles in Maryland and Virginia; and local fire vehicles. Although less well developed than the National Model, the CapWIN can shed light on issues related to information integration and information sharing among multi-state agencies.

6.2.3    Other Automated Data Recording Technologies

            There are a number of devices specifically deployed to improve highway safety such as on-board personal computers, on-board event data recorders, and red light running cameras. Although little or none of the data collected from these devices are archived and shared due to concerns about privacy violation and liability, the potential of these technologies and their archived data are being actively discussed.

            Information on the vehicles that are involved in highway crashes can be currently extracted from different sources such as the police’s crash reports and the event data recorders (black boxes) installed in vehicles. In fact, data collected from event data recorders not only provide vehicle information but also information on the driver’s reaction, from pre-crash to post-crash periods.

Event Data Recorders (EDR)

            The National Transportation Safety Board (NTSB) encouraged automobile manufacturers and the National Highway Traffic Safety Administration (NHTSA) to work together to collect crash data using on-board collision sensing and recording devices. General Motors (GM) began using event data recorders with the introduction of the air-bag in 1974, which collected information on the air-bag status and crash severity. Since that time, the event data recorders used by GM have evolved and allowed the recording of a greater number of data elements. GM’s 1999 event data recorder or Sensing & Diagnostic Module (SDM) is capable of recording the following information:

            ●    State of the warning indicator when the event occurred (ON/OFF),

            ●    Length of time the warning lamp was illuminated,

            ●    Crash-sensing activation times or sensing criteria met,

            ●    Time from vehicle impact to deployment,

            ●    Diagnostic trouble codes present at the time of the event,

            ●    Ignition cycle count at event time,

            ●    Maximum delta V (change in longitudinal vehicle velocity) for near-deployment,

            ●    Delta V vs. time for front air-bag deployment event,

            ●    Time from vehicle impact to time of maximum delta V,

            ●    State of driver’s seat belt switch,

            ●    Time between near-deploy and deploy event (if within 5 seconds),

            ●    Passenger’s air-bag enabled or disabled state,

            ●    Engine speed (5 seconds before impact in 1 second intervals),

            ●    Vehicle speed (5 seconds before impact in 1 second intervals),

            ●    Brake status (5 seconds before impact in 1 second intervals), and

            ●    Throttle position (5 seconds before impact in 1 second intervals).

Once an air-bag deployment is recorded, this information is permanently stored and cannot be altered, erased or cleared by service or crash investigation personnel.

            General Motors installs some type of EDR in most of its vehicles, and is the only manufacturer that shares its crash data with the research community. Other OEMs (e.g., Ford, Honda, etc.) also began to install event data recorder modules in some of their 2000 model year vehicles, although Ford collects fewer data elements. According to a report titled: “Recording Automotive Crash Event Data,” NHTSA has begun to build a database of crashed GM vehicles by retrieving data from the event data recorders of the wrecked cars.^6.5

            It is anticipated that crash data provided by the onboard event data recorders could provide NHTSA and automotive engineers with a wealth of real-world crash data (crash pulses) which can never be obtained in a laboratory setting. These data would allow manufacturers to produce more crash-worthy vehicles, reducing the total number of injuries and fatalities. In 1992, GM installed crash data recorders on 70 Indianapolis 500 race cars. The data collected provided a better understanding of impact tolerances of human beings, which in turn resulted in design changes of the race cars. These design changes are believed to have substantially reduced the number of serious driver injuries during the 1998 racing season.

            It is also anticipated that, aside from improving the crash worthiness of vehicles, the data collected can benefit the design of highway infrastructure. Materials as well as the design of guardrails and crash barriers can be improved to better absorb vehicle impacts while other roadside components such as light poles and road signs can be improved to minimize injury in the event of an impact.

            Using archived data from the event data recorders to achieve the greatest safety benefits, one has to first address the issues of data ownership, privacy and liability. GM maintains that any data recorded by on-board sensing devices belong to the vehicle owner and is private information. It has been suggested that experts in the aviation community be consulted when data ownership concerns arise.

            Aside from the data ownership concern, it was reported that EDR data alone does not provide third party researchers with enough information to make it particularly useful.^6.6 Other crash information such as accident reports and photographs are essential to give the data needed context and meaning. But, the format of this information is not compatible with the EDR data. It was recommended by this National Cooperative Highway Research Program (NCHRP) effort that NHTSA considers the feasibility of summarizing EDR data in such a way that they would be useful to third party researchers.

            National Highway Traffic Safety Administration is embarking on a research project in year 2001 where specially designed data recorders will be installed in a large fleet of “participating” vehicles to collect a wealth of information on vehicle movements, the type of roadway on which the vehicle is traveling, velocity, seat belt use, time stamp, etc.^6.7 This information will be integrated with traffic data collected from TMCs to better understand crash propensity in actual traffic stream such as the crash impact of lane weaving under different traffic conditions.

Red Light Running Cameras

            Each year, more than 1.8 million intersection crashes occur. In 1998, red-light-running (RLR) crashes accounted for 89,000 crashes, 80,000 injuries and nearly 1,000 deaths. Public costs exceed 7 billion.^6.8 To raise awareness of the dangers of red light running and to reduce fatalities, FHWA created the Stop Red Light Running Program in 1995.

            New technologies are used to help enforce RLR violations. Typically, a photo detection system is installed at an intersection. When the traffic signal turns red, the system becomes active and the camera takes pictures when vehicles enter the intersection. The camera records the date, time of day, time elapsed since beginning of the red traffic signal, and the speed of the vehicle. Camera films are loaded, unloaded and processed on a regular basis.

            A 1999 FHWA report synthesizes the results of a number of RLR demonstrations around the country.^6.9 Data collected from the photo enforcement cameras were used to evaluate the effectiveness of this approach. Almost all demonstration sites experienced significant reduction in the number of traffic violations. The FHWA report concluded that at least a 20% and as much as a 60% reduction in RLR violations could result from the implementation of an electronic enforcement program. Although Howard County, Maryland reported a reduction in the number of crashes at the instrumented intersections, this conclusion was not based on data archived/stored in the cameras.

6.3 BARRIERS TO ARCHIVE, SHARE AND ANALYZE DATA PERTINENT TO HIGHWAY SAFETY

6.3.1 Privacy Issues and Barriers

            Video surveillance cameras can improve response time of emergency vehicles to the scene of an accident and, Red Light Running (RLR) camera systems can significantly reduce the number of accidents occurring at intersections with video-enforced traffic lights.^6.10 Although these technologies are supported by courts when used by traffic management centers or law enforcement, it is unclear what attitudes will be regarding the sharing of this data with unrelated third parties.

            The perceived benefits of, and barriers to, sharing video data were surveyed among stakeholders of TransGuide in San Antonio.^6.11 The issue of sharing was about the possible use of TransGuide’s ATMS video data by public agencies and stakeholders in San Antonio other than the Texas Department of Transportation (TxDOT) which has proprietary control of the data. The biggest public sector beneficiary of video sharing is the San Antonio Police Department (SAPD). Using video, the SAPD can verify an incident report without dispatching a patrol unit, and can respond to incidents within two minutes of detection. TransGuide video data are typically not recorded and archived unless it is explicitly requested The “sharing” component of TransGuide largely involves the distribution of ATMS-related data over the Internet, television, and FM radio. In TransGuide, sharing also implies informal arrangements between TxDOT and other public agencies such as the San Antonio Fire Department and Emergency Management Service. The agreement allows stakeholders to have access to live video feeds and traffic data, and in some cases allows limited control over the camera equipment.

            Based on the results from the ITS Deployment Tracking Surveys, the obstacles to sharing data are: concerns about having a bureaucratic process, and privacy issues. Privacy issues center on police access to video data, and concerns over how the data would be used, especially for law enforcement purposes. The SAPD’s response to these concerns is that it does not use TransGuide video data for law enforcement, and has no plans to do so.

            Although the privacy concern varies from one agency to the next, there is a clear need to investigate the use of privacy protection technologies, such as cryptographic technologies, to remove these concerns. A balance between the right to information and the protection of personal data should be struck when implementing these technologies. Finally, any ADUS activity should adhere to guidelines that may limit government access to communications and stored digital information. Examples of such rulings include the Electronic Rights for the 21st Century Act, and the European Union's Data Protection Directive. Approaches to overcome liability and privacy concerns are beyond the scope of this work and will not be discussed in detail.

            Although cell phones may not strictly qualify as an ITS technology, they are designed to be mobile and are often used during transportation. There are about 83,000 wireless calls to 911 per day. In addition to their obvious safety use for calling 911, cell phones also produce electronic data.^6.12 Although cell phones are designed for direct vocal communication, the electronic data they produce can also have an indirect, positive influence on safety. They emit signals that can be located in the event of an emergency call either by triangulation or via a GPS chip embedded in the phone. However, using a cell phone as a locator device comes with privacy concerns. Although it should be noted that privacy is a concern here, it does not appear to be a barrier in the context of 911 emergencies. Ordinary land line telephones provide 911 operators with location information; therefore, it would be hard to make the argument that determining the location of a 911 call from a wireless phone for emergency purposes constitutes an unreasonable invasion of privacy. Furthermore, in April 1998, the Federal Communications Commission ordered cell phone providers to provide 911 operators with the address of the cell phone tower sending the distress call.

            Event data recorders placed in automobiles are another type of recording device that raises concerns about privacy and liability. “Some people argue that black boxes in cars are a violation of privacy, but the American Coalition for Traffic Safety disagrees. It says the systems are derived from sensors and computer modules, do NOT include voice or video recorders, and thus are not a violation of privacy.”^6.13 This does not ensure that the use of event data recorders are or will be free from litigation, only that invasion of privacy will probably not be the most serious barrier to the use of this technology.

            Data ownership, however, could prove to be a delicate and complex issue for the use of event data recorders. Does the owner of the car also own the data contained in the data recorder, or does the manufacturer? Do law enforcement officials have a right to that data or do they need permission or a warrant? There are also many groups who are interested in crash data such as insurance companies, courts, NHTSA, the media, manufacturers of safety equipment and maybe even those with less than noble intentions. Clear guidelines about access and ownership of in-vehicle event-data recorders will have to be established before this type of data can be widely used. Similar guidelines can be established for sharing other ITS data such as data collected from red-light running cameras or incident recorders at the intersections.

6.3.2    Technical and Technological Issues and Barriers

            Technology is a significant barrier for highway safety applications of ADUS data. Event data recorders are still relatively new in automobiles and must accurately record automotive systems under the extreme conditions of an impact. Data reliability of event data recorders will have to be studied before this data can be used with confidence in courts and by other disciplines where accuracy is absolutely essential. Additionally, different manufacturers are using different types of event data recorders which can result in problems for data integration.

            Although collection of location data from cell phones has been mandated for 911 calls, it is not yet completely implemented. By October 2001, cellular carriers were required to be able to locate callers with much greater accuracy. There may be several different ways to comply with this new regulation which may provide data with varying degrees of accuracy or in varying formats which would complicate the use of such data.

            That said, the biggest challenge in using ADUS for highway safety applications is probably the extraction of information from the archived images in a form that is suitable for analysis. Much of the information that is currently archived, and that can be used to address intersection safety and speed management, is in image format. For example, images that are, or could be, archived from red-light enforcement cameras at intersections are typically stored in image formats (Figure 6.8). Manual processing of these images can be extremely labor intensive and resource demanding. The transformation of such images, into machine-readable form for analyses, is a monumental challenge to the user of this information. Feature detection and/or optical character recognition (OCR) have the potential to address these challenges. However, the feasibility of these technologies needs to be adequately tested and their relative costs and benefits assessed.

Figure 6.8 Archived Images from Red Light Enforcement Camera

Source: http://www.city.winnipeg.mb.ca/police/Traffic/photo%20Enforcement/redlightexample.htm

6.3.3    Institutional Issues and Other Barriers

            The institutional and other barriers associated with implementing ADUS are largely the same among the different ADUS applications. These common barriers such as cost, proprietary rights, politics, and a lack of understanding or knowledge about ADUS are covered in further detail in Chapter 2. Aside from these common barriers, the implementation of ADUS, as it relates to highway safety, faces another problem. Of all the data elements collected by ITS deployments, none of them collect data that are strictly about safety. Traffic management centers collect data in order to keep traffic flowing smoothly, which may benefit safety, but safety remains tangential to their work. Those who are interested in safety have to find their own ways to relate ITS-generated data to the topic of highway safety.

            Another safety issue is the heightened concern over protecting the identity of incident victims. This heightened sensitivity can slow down or limit the flow of certain types of data. Apart from the barriers common to data archiving in general, archiving highway safety data is confronted by a very different challenge. Part of the traffic information on freeways and arterial roads is collected by cameras. In particular, video cameras are used to identify the exact locations and circumstances where something affects traffic on the freeway system. That type of information is enormously valuable in advancing our understanding of crash causation so that effective countermeasures can be deployed. Notwithstanding, access to that information has been extremely controversial. Undoubtedly, the biggest barriers to archiving data that could have some bearing on safety applications are privacy and liability issues. Recognizing these concerns, many agencies limit their data archiving or withhold safety information and camera images. On limited occasions, images are recorded for traffic studies (such as vehicle counts, weaving movements), training purposes, and exceptional circumstances (such as in criminal investigations).^6.14

6.4 POTENTIAL FOR HIGHWAY SAFETY APPLICATIONS

            ITS technologies such as event data recorders in vehicles and RLR cameras at signalized intersections have a direct influence on highway safety. Archived data from event data recorders provide accurate crash data leading to better vehicle and highway infrastructure designs and regulations while RLR cameras encourage motorists to respect red lights to avoid traffic citations and points on their driving record thus reducing the number of traffic accidents and fatalities.

            Other technologies that are more on the fringe of ITS such as onboard police computer systems and cell phones can also impact safety though in a more indirect way. While onboard police computer systems are primarily designed to improve the efficiency of police reporting and limit the introduction of errors while handling evidence, they provide a secondary benefit that improves safety. By reducing the amount of manual paperwork required of officers and allowing them to work directly from their police cruisers, the officers are able to spend more of their time on patrol, where they can respond to incidents more quickly or serve as a deterrent to criminal or reckless activity.

            The Criminal Investigative and Traffic Safety Incident Information Management System, or IIMS, is a program that is being implemented in Pennsylvania in order to give the Pennsylvania State Police an edge in the enforcement of crime and traffic safety. This system seeks to make every police cruiser in Pennsylvania a mobile office, providing the officers with all the information they need through onboard computer access to state and federal databases. It is estimated that this system will reduce the time officers spend on paperwork by more than half, which would have the effect of doubling the size of the police force on the street. This system is expected to be fully operational within 3 years and will include the following elements:

            ●    Onboard computer allowing access to state and federal databases.

            ●    Interface allowing officers to enter their report information directly into the system at the scene of the incident.

            ●    Onboard GIS system to help officers locate an incident more quickly.

            ●    Bar code evidence handling system.

            ●    Portable fingerprint device linked to state and federal records for rapid identification.

            ●    Information entered into the system by the police will be shared with other agencies.^6.15

            The most effective data collection undertaking is driven by information needs. Without any specific application, any attempt to highlight the potential of archived data to address highway safety concerns is complicated by the complexity of crash occurrences. Nonetheless, we examined individual ITS infrastructure components separately and identified how data, if collected and archived, could be used to address safety concerns. Table 6.4 offers a list of potential uses of data archived from roadway systems, which might have some bearing on highway safety applications. From the perspective of the Intelligent Vehicle Initiative (IVI) Program, there are four Generation of Field Operational Tests (FOT) that hold promise in data archiving. These FOTs involve the use of crash avoidance technologies to improve highway safety. They are:

            ●    testing the operational effectiveness of the Rollover Stability Advisor (RSA), Rollover Stability Controller (RSC), and Lane Tracker (LT).

            ●    testing the operational effectiveness of the Infrastructure-Assisted Hazard Warning (IAHW) system and Automatic Collision Notification (ACN).

            ●    testing the operational effectiveness of the bundled advanced safety system of Electronically Controlled Brake Systems (EBS), Collision Warning Systems (CWS), and Adaptive Cruise Control (ACC).

            ●    testing driver assisting technologies that include magnetic roadway tapes, DGPS, GIS Mapping, 360 degree obstacle detection devices, forward Collision Avoidance Systems (CAS), Head-Up Displays (HUD), auditory warnings, external light warning systems, in-vehicle proximity warning systems, and Micro-Data Acquisition Systems (Micro-DAS). Although the primary focus has been on snow plow vehicles thus far, ambulances and police vehicles will also be included.

The integration in FOTs of data archived from these in-vehicle technologies, geometric data on roadways, and data on vehicle characteristics and functions and weather conditions will accelerate the understanding of the complex interrelationships between vehicles, roadways and environments on the occurrence of truck crashes.

            Safety analysis and planning have been afflicted by incomplete, inconsistent and not-so-timely information. One system that attempts to overcome this impediment is particularly noteworthy. The Advanced Law Enforcement & Response Technology (ALERT) initiative has equipped about 10 police vehicles (a few at the College Station Police Department in Texas and a few for the Secret Police around the White House) with on-board and hand-held computers and other devices. One feature of the system is to streamline data collection and sharing, and improve communication between law enforcement and the first-response community. Unfortunately, ALERT will no longer be implemented largely due to the lack of funding resources.

            In order to meet the safety objective in U.S. Department of Transportation’s strategic plan, the safety considerations are to be integrated in the transportation planning processes at all levels of government, specifically in the Statewide Transportation Improvement Programs and the Transportation Improvement Programs developed by state DOTs and MPOs, respectively.^6.16 In a recent workshop hosted by the National Association of Governor’s Highway Safety Representatives, good quality data and robust analysis were identified amongst the strategies important to successfully integrating safety into the transportation planning process. ITS-generated data hold promise in identifying and defining high-risk areas, and predicting the probability of crash and incident occurrences, and providing the information needed for developing and implementing effective measures for improving safety.

6.5 OPPORTUNITIES

       There are many ways that archived ITS data could improve highway safety. Rather than try to identify all possible opportunities, this section identifies those that appear to be feasible, can be quickly deployed, and are most likely to produce immediate benefits/results. The rationale for identifying these “low-hanging fruits” is that the sooner that quantifiable benefits of using ITS-generated data for safety improvement are demonstrated and disseminated, the sooner additional deployments will be stimulated. Any of these opportunities identified below can be developed into a FOT with public and private partnership. 

6.5.1 ADUS for Intersection Safety

       More than 2.8 million crashes occur annually at intersections, killing 10,000 people and injuring another 1.5 million people. These crashes account for 45% of all reported crashes. As a result, the Federal Highway Administration has identified intersection safety as one of the four high-risk areas; and the American Association of State Highway and Transportation Officials highlighted the need to improve the design and operation of our highway intersections in its Strategic Highway Safety Plan.

       Participants in a recent workshop^6.17 recognized the need to implement proven, effective intersection safety technologies (e.g., red light running enforcement camera, signal timing), and to develop data and tools to analyze safety at intersections. One of the actions needed to meet these priorities is the improvement of databases on traffic, roadway, signal timing, and crash.

       Images collected from enforcement cameras, and archived data on weather and roadway conditions, traffic volume, and vehicle turning movements can facilitate the identification of the reasons for the most common and severe types of crashes at intersections, and an understanding of the inter-relationships among these factors. The integration of these data can also be used to evaluate the effectiveness of different intersection safety technologies.

       With the infrequency of crash occurrence, one of the challenges in crash analysis has always been small sample sizes. To overcome this challenge, any efforts to use these archived data for intersection crash analysis needs to include data collected/archived from enforcement cameras around the country. A carefully designed plan to harvest the archived data/images is the essential first-step.

6.5.2 ADUS for Work Zone Safety

            Work zones in U.S. highways have become increasingly dangerous for both workers and travelers. An estimate for the year 2000 was that there were about two deaths per day from work-zone related accidents.^6.18 Overall, approximately 800 people are killed, and forty thousand people are injured every year as a result of motor vehicle crashes in work zones. The safe and efficient flow of traffic through work zones has been, and continues to be, a major concern to the transportation community.

            Decision-making tools can be effective in improving work zone safety through planning, traffic management, traveler information, and traffic control. Motorists can be alerted to traffic congestion by portable message boards and a highway advisory radio system. An effort is underway by the FHWA to develop user friendly PC-based decision making tools that will accurately analyze and reliably predict work zone impacts. These tools will allow practitioners involved in the project pre-planning, planning, development, construction and maintenance phases to weigh alternative strategies to mitigate the mobility and safety impacts of work zones. ADUS should be an integrated and an essential component of these tools.

            Decision making tools become useful only when decisions are made based on realistic and current data. Oftentimes, decisions are made based solely on subjective judgements or “similar” data. If archived data are used to develop trends and patterns, decisions can be made that are more accurate for a specific locale or facility.

            Furthermore, archived information from variable electronic message boards, that inform motorists with real-time traffic delay and alternative routes, can be integrated with archived data on traffic and speed that were recorded before and after the message was broadcasted. This confirmation of data can help evaluate the safety impacts of alternative traffic management plans.

 6.5.3 ADUS for Speed Management

            According to the Fatal Accident Reporting System, almost 1 of every 3 traffic fatalities is related to speeding. These crashes also cost the American economy approximately $27.7 billion each year.

            With the repeal of the National Maximum Speed Limit, there is renewed interest in how best to set speed limits. Speed management is a complex issue involving engineering, driver behavior, education and enforcement. Research is underway to develop and test strategies and technologies to manage speeds and encourage wider adoption of speeds appropriate for the particular class of road, roadway design, and travel conditions.

            Travel speed has been the most commonly collected and archived data element from the Freeway and Arterial Management Systems. These data are most appropriate for establishing the relationship between travel conditions and speeds, analyzing the safety implications of different speeds, and examining the feasibility of speed limits that are adapted to current travel and roadway conditions.

6.5.4 ADUS for Pedestrian and Bicycle Safety

            Six thousand pedestrians are killed and another ninety thousand pedestrians are injured every year in this country. It is estimated that pedestrian injuries and fatalities have resulted in $20 billion in societal costs every year.^6.19

            Analysis of pedestrian and bicycle safety has been hindered by the lack of data. Images from cameras installed at signalized intersections can help decipher the events that lead up to a crash. That said, a feasibility study should be conducted before a full-scale study is implemented.

            Several technologies (such as microwave pedestrian detection) are being deployed to minimize vehicle and pedestrian conflicts. The integration among these technologies, AVL in transit vehicles, and adaptive signal timing might further reduce these conflicts. This notion is similar to the current use of signal over-ride for transit vehicles and emergency vehicles on arterials. The feasibility of this idea can be examined by analyzing the archived data from cameras installed at pedestrian crossings, and archived data from signal timing plans. Again, a carefully designed plan to harvest the archived data/images is the essential first-step.

6.5.5 ADUS for Highway Safety Public Awareness and Training

            Archived images from surveillance cameras installed in construction and maintenance areas or at intersections can be used to increase public safety awareness. However, a comprehensive testing of privacy-protection technologies should be the first step.

            As previously mentioned, any of these opportunities can be developed into an FOT with the goals of:

            ●    identifying technical and institutional barriers to archiving, using, and sharing ITS-generated data;

            ●    developing solutions to overcome these barriers;

            ●    identifying issues pertinent to standards development;

            ●    examining the feasibility of integrating ITS-generated data with data collected from traditional and emerging technologies (e.g., highway monitoring data, remotely sensed data);

            ●    identifying and quantifying costs and benefits;

            ●    disseminating lessons learned, and

            ●    sharing the developed procedures and software in an open-source environment. Some examples of these procedures and software are: those developed to convert raw ITS-generated data into formats acceptable to existing and/or off-the-shelf data management or analysis software, check the quality of the data, impute missing data, correct questionable data, abstract information suitable for data analysis from “text” files, estimate potential recurring and non-recurring traffic delays, and other applications. The benefit of sharing these procedures and software in an open-source environment is that it reduces the “re-inventing the wheel” thus enabling more efficient use of resources.

ENDNOTES:

Return to Table of Contents