《SHRP-II-Request-for-Proposals-S05.pdf》由会员分享，可在线阅读，更多相关《SHRP-II-Request-for-Proposals-S05.pdf（24页珍藏版）》请在三一文库上搜索。
1、 - 1 - SHRP II Request for Proposals Safety Focus Area Project Number: S05 Project Title: Design of the In-Vehicle Driving Behavior and Crash Risk Study Date posted: September 13, 2006 SHRP II BACKGROUND To address the challenges of moving people and goods efficiently and safely on the nations highw
2、ays, Congress has created the second Strategic Highway Research Program (SHRP II). SHRP II is a targeted, short-term research program carried out through competitively awarded contracts to qualified researchers in the academic, private, and public sectors. SHRP II addresses four strategic focus area
3、s: the role of human behavior in highway safety (Safety); rapid highway renewal (Renewal); congestion reduction through improved travel time reliability (Reliability); and transportation planning that better integrates community, economic, and environmental considerations into new highway capacity (
4、Capacity). Under current legislative provisions, SHRP II will receive approximately $150 million with a total program duration of seven years. Additional information about SHRP II can be found on the programs web site at www.trb.org/shrpii. OVERVIEW OF SAFETY FOCUS AREA The focus of the SHRP II safe
5、ty plan is a large field study of driving behavior using volunteer drivers and a sophisticated instrumentation package installed in the volunteers vehicles for at least several months but not more than one year. This SHRP II field study is intended to support a comprehensive safety assessment of how
6、 driver behavior and performance interact with roadway, environmental, vehicular, and human factors and the influence of these factors and their interactions on collision risk, especially lane departure and intersection collisions. The accompanying chart lays out the main projects anticipated for SH
7、RP II Safety research. The chart provides a general idea of the flow of work and of the placement of Project S05, for which this request for proposals (RFP) has been issued. The exact number, content, and timing of contracts are subject to change. Project S05, the Study Design, includes the developm
8、ent of a complete data collection system, a field trial of the system, and the management plan for the full study. The Study Design is supported by Project S03 to develop a digital measurement van to collect detailed and accurate roadway information in the study areas and by Project S01 to identify
9、analytic methods to address the research questions. Multiple analysis development projects are anticipated under Project S01. The full-scale in-vehicle field study will be implemented at the beginning of the third year in 3 to 4 study areas with an overall quality assurance and technical coordinatio
10、n contractor. The total anticipated budget for the field study is $30,000,000. Roadway data will be collected in the study areas under a separate project, S04. Multiple analysis projects (S08) are planned to address a wide range of research questions using the data collected. - 2 - In-Vehicle Study
11、Roadway Data Analysis Site-Based Study 2011 S05 (Phase II): Technical Coordination and Independent Quality Assurance for Field Study S05 (Phase I): Design of the In- Vehicle Driving Behavior and Crash Risk Study 20092010 Prepare for field study 2008 S02: Integrate Methods and Develop Analysis Plan S
12、01: Development of Analysis Methods Using Recent Data (multiple awards, two phases) Study Design the exact dates will be coordinated through SHRP II staff. Information from the Analysis Methods contractors may be used by the Study Design contractor to specify the data requirements for the instrument
13、ation packages to be used in the field study. Data requirements must be specified early in the study design process, which is why this workshop is called for early in the period of performance of all three projects. The objective of the workshop is to exchange information on implications of the anal
14、ysis methods under consideration for data elements to be collected, data storage, and data access provisions. Workshop participants will include the Study Design contractor, the contractors from Projects S01 and S09, SHRP II Safety Technical Coordination Committee members, SHRP II staff, - 6 - and p
15、ossibly other experts. The only travel expenses for which the Study Design Contractor will be responsible are those of the Study Design contractors own staff. Products: (1) An overall plan of the analytic methods to be used in addressing the research questions of Task 1, in enough detail to provide
16、the basis for making decisions on the sampling plan and the data to be acquired (which drivers? which vehicles? how much data are needed statistical power? what exposure data are needed? how should exposure data be selected independence? etc.). (2) A workshop with the contractors on Projects S01 and
17、 S09 to discuss analytic approaches and the implications for data collection, data storage and data retrieval. Task 3: Sample Design The objective of this task is to produce a sample design for the full-scale in-vehicle field study. Subtasks: a. Determine the driver and vehicle sample best suited to
18、 answer the research questions of Task 1 using the analysis methods of Task 2. The sample should reflect established findings about the interaction of driver age, gender, and vehicle type in exposure-based analyses of crash involvements. A baseline/control subgroup described by age X gender X vehicl
19、e types that are not overrepresented in crashes or fatal crashes should also be included in the sample. The sampling plan should allow the flexibility to change the driver-vehicle mix when the instrumentation package is moved to different vehicles. At the planning stage, it is believed that a minimu
20、m of three age groups are needed, to separate young (e.g., 16-19), adult (e.g., 30-45) and older (e.g., 70+) drivers. If resources permit, a middle-aged driver cohort (e.g., 55-69) may be useful. Vehicles are limited to passenger vehicles; motorcycles, large trucks, and other specialized vehicles ar
21、e not included. Possible vehicle groups are small, medium, and large passenger cars, vans, pick-up trucks, and sport utility vehicles. The number of subgroups may have to be reduced to achieve reasonable minimum sample sizes in each, given the total sample. Registration data or insurance information
22、 may reveal if there are adequate populations for all the targeted combinations of age, gender, and vehicle type. Not all age and gender drivers may be represented in each vehicle type. Subgroups that do not have adequate representation in the population data might be omitted from the sample design.
23、 b. Determine the geographical areas in which the study will be conducted. For planning purposes, the study should consider being conducted in 3 or 4 geographic areas. Geographic area selection must include a mix of roadway and driving characteristics - rural, suburban, and urban; Interstate-type ro
24、adways and local streets; different weather conditions. Geographic area selection also must balance the need for geographic diversity with the goal of minimizing the costs of the field study. In selecting geographical sampling areas, consider using NHTSAs NASS sampling areas, to provide opportunitie
25、s for in-depth crash investigation and for more efficient linking of study data with national data sets such as NASS and NMVCCS. Consider also - 7 - the availability and adequacy of related data: road inventory, network GIS, state crash reports and driver records, insurance, etc. Product: A sample d
26、esign that addresses the selection of study areas and selection of drivers and vehicles within these areas. Task 4: Sample Design Interim Report DELIVERABLE: Prepare a Sample Design Interim Report detailing the activities of Tasks 1 through 3, including all the products called for under these tasksr
27、esearch questions, analysis plan, sample plan. Part II: Develop the Complete Data System Task 5: Driver Testing The objective of this task is to identify and/or develop a suite of tests for volunteer drivers in the field study. A central focus of the in-vehicle driving study is driver behavior. An i
28、mportant group of variables that may serve as independent or predictor variables in the planned analyses will come from comprehensive testing and monitoring of the drivers participating in the study. The goal is to minimize the duration of driver testing, which should not exceed two (2) hours per pe
29、rson. Subtasks: a. Identify and/or develop measurement tools to acquire all driver characteristics desired at the beginning of study involvement; and select a preferred knowledge test instrument. b. Identify a practical means of updating changes in drivers medical conditions and/or medication use th
30、at occur during their months of study involvement. The following requirements should be reflected in the selected suite of driver tests: Testing of driver functional abilities needed to drive safely will be conducted before each driver enters the study; some of these tests may be repeated when the d
31、river leaves the study, to measure any changes. Drivers preexisting medical conditions and use of medications, if any, will be documented through self-report; a questionnaire will be administered at the beginning of each persons study involvement, and a means will be provided for him/her to report c
32、hanges that occur during the study. Measures of driver functional ability should include selected aspects of vision, hearing, cognitive ability, and physical fitness/impairment. The selection of the medical conditions and medication usages targeted for self-report should reflect the latest findings
33、in this active area of research. In all cases, priority should be given to measures for which a significant relationship with crashes (preferably at-fault crashes) has been demonstrated. The same information should be obtained for all participating drivers, regardless of age. The study also should a
34、ccess driver history information from the states in which the study is conducted. This will include traffic convictions and crashes, plus - 8 - suspensions/revocations and any other administrative actions, within the preceding 3 years (or since the time of first licensure if less than 3 years). Driv
35、er history information will be collected for each driver at the beginning and end of study involvement. Measures of safe driving knowledge should be obtained for each participating driver when he/she enters the study. This should be accomplished using a common instrumenteither a current state knowle
36、dge test for beginning drivers, or a recently developed and widely recognized model test. Driver testing will not be focused on transient states (e.g., fatigue, anger, distraction) that may be inferred from the on-board video and other data elements to be collected in the study. Product: A suite of
37、existing or new procedures and techniques to characterize the drivers in the study. Task 6: Driver Face and Other Video Recording and Processing The objective of this task is to address the issues associated with the use of video data (primarily the driver face video) in the full-scale field study a
38、nd to build, program, and test all necessary hardware and software. Video information will be critical for many analysis questions. Currently, it is very time consuming for an analyst (usually multiple analysts for reliability) to review video data in order to extract information. Some work has been
39、 done to automate aspects of this process. Subtasks: a. Identify the types of driver face video information that are likely to be of interest (e.g., drowsiness, distraction, etc.) and specific measures that provide those data. Assess what can currently be collected and how. Review and summarize the
40、potentially useful measures for this project that could be obtained from such analyses (refer to SAE, ISO documents and research that has been carried out, including the NHTSA 100 car study, various European Union projects etc). Based on the research questions developed under Task 1, provide researc
41、h-based justification for the choice of dependent variables to address those questions. Include recent large-scale studies in your review. Assess advantages and disadvantages of alternatives. b. Review and assess the current state-of-the-art for post-processing of video (not restricted to the drivin
42、g domain). Provide a summary of what can be accomplished through manual decomposition and what can be done with automated methods. Appraise the status of any current automated video analysis to address glance direction/duration, and the PERCLOS fatigue measure. c. Explore development of new approach
43、es to automated processing of the driver face video. Report on current status and feasibility. What gaps need to be addressed and how could this work be accomplished? Assess the value-added and the pros and cons of this work. Include work within the driving domain such as the NHTSA-sponsored work on
44、 automated approaches to video analyses. If existing software is not adequate, this project should undertake development of such automated video processing, if feasible within project resources and approved by SHRP II. - 9 - d. Select an approach for the driver face video processing. Consider feasib
45、ility, cost, and benefits. What are the tradeoffs among the various approaches in terms of data reliability, cost, ability to go back to answer further questions later, etc. Consider multi- level types of approaches where more general questions may be answered by a coarser analysis of much of the da
46、ta and more targeted analyses could be made for data subsets of interest. e. Identify other sorts of visual information (non-driver video) that are desirable, such as forward vehicle view, backward view, camera views that would assess ambient traffic, etc. f. Specify all video measurements and metho
47、ds. g. Specify equipment, procedures, data collection, processing, and data reduction methods. h. Illustrate the proof of concept: build, test, and evaluate. Products: (1) Review of current state-of-the-art, with annotated bibliography of papers. (2) Specifications for collection, automation, and re
48、duction of driver face and other video information for use in analyses. (3) Fully functional prototype hardware and software. Task 7: Data Items and Instrumentation Package Specifications The objective of this task is to address all aspects of the specifications, production and testing of a data col
49、lection instrumentation package for the study vehicles. Several recent studies, including NHTSAs Naturalistic Driving Study, have developed instrumentation packages for passenger cars. It is expected that a custom data-collection package will need to be fabricated using state-of-the-art equipment to meet the study needs. Suitable packages are not currently available commercially. The SHRP II field study envisions a minimum of 2,500 hardware packages to be re-installed in perhaps three or four times that many vehicles over 2-3 years. The anticip