Freight Transportation Surveys (2011)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

12 CHAPTER THREE FREIGHT TRANSPORTATION SURVEYS: STATE OF THE PRACTICE INTRODUCTION This chapter describes the state of the practice as determined from the practitioner surveys. Following this introduction, the discussion is organized according to the six survey sections: Applications of existing data collection are described 1. in Applications. Survey methods and data characteristics are described 2. in State of the Practice for each of the different types of surveys that are included in the study scope. The Survey Costs section discusses survey costs for each of the survey types. The Data Availability and Dis- semination section reviews data availability and dissemination. Practitioners’ freight data requirements are discussed 3. in Freight Data Requirements. Practitioners’ use of public and commercial data 4. sources is presented in Use of Existing Data Sets. Practitioners’ use of ITS technologies for surveys 5. and data collection is described in Use of Intelligent Transportation System Technologies. Finally, User Assessment of Data presents practitio-6. ners’ assessment of how well current surveys and data meet their needs. The section also provides a discus- sion of lessons learned. It is important to note that the information was provided in confidence. Accordingly, individual respondents or facili- ties are identified only generally (e.g., “state DOT”). Appen- dix A provides a complete summary of the survey results and is a web-only document. APPLICATIONS Table 2 tabulates the purposes and transportation modes that respondents consider in their freight transportation surveys and/or data assembly. Note that respondents were asked to identify all relevant applications and modes. TABLE 2 APPLICATIONS AND MODES CONSIDERED IN FREIGHT TRANSPORTATION SURVEYS/DATA Application/ Mode Truck/ Highway Rail Air Marine Intermodal/ Cross- modal/ Multimodal Total Infrastructure Capacity Planning 43 26 19 19 27 134 Modeling 39 11 7 6 9 72 Cost-benefit Analysis 24 17 7 10 12 70 Land Use Planning 20 6 8 8 9 51 Operations and/or Safety Analysis 32 17 14 10 15 88 Environmen- tal Impacts 25 16 14 12 16 83 Policy 38 28 24 25 29 144 Total 221 121 93 90 117 642 Several observations may be made regarding this table: The range of applications was broad, with 642 applica-• tions cited. The dominant applications were policy (144 citations) and infrastructure capacity planning (134). Modeling was well down on the list, at 72 citations; oper- ations/safety analysis (88) and environmental impacts (83) each garnered a greater number of citations. All types of applications were cited, with cost-benefit anal- ysis cited almost as frequently as modeling (70 times) and land use planning cited the fewest (51) times. This broader range is consistent with the findings of a 2004 study of urban freight data needs, which pointed out that the data are needed for infrastructure planning, opera- tions, safety, and environmental issues, in addition to the more traditional applications in planning and modeling. That study also grouped data needs into five categories: cargo; road transportation; major freight generators and corridors; non-road transportation modes; and eco- nomic, land use, and socioeconomic data (1). Trucks/highways were the dominant mode, at 221 • citations. However, each of the other modes also was important. Rail was next, at 121 citations. Intermodal, cross-modal, and multi-modal had 117 citations. Next

13 ing by seven respondents (the same number that cited manual license plate matching), and Global Positioning System (GPS) vehicle tracking by five respondents. Commercial vehicle trip diaries were cited only by four respondents. Other survey types comprised [truck] toll revenues, motor carrier identifi- cation surveys, and multimodal freight studies. FIGURE 1 Types of surveys conducted by respondents. Table 3 compares the applications for which the surveys were (or were to be) used. Respondents were asked to indicate all applications, and the number of citations—306—indicated that the surveys commonly were used for many applications. By far the most common application was infrastructure or facility planning, with 88 of 296 citations (30%). Demand management and traffic safety applications were next, at 52 and 49 citations respectively. Logistics planning, land use planning, and air quality management followed, at 34, 26, and 24 citations respectively. These applications generally were well distributed across all survey types. Several other pur- poses also were cited, although never exceeding five citations: notably, modeling was cited by only five respondents. Table 4 indicates that many respondents conducted sev- eral types of surveys and data activities jointly. Roadside/ intercept surveys (40 citations), telephone surveys (20), mail- back surveys (18), personal interviews (26), and focus group/ stakeholder surveys (18) most commonly were conducted with other survey types, notably personal interviews (31), telephone surveys (26), and focus group/stakeholder surveys (25). Table 5 describes the geographic coverage of the surveys. The coverage ranged from facility- and corridor-specific to urban, state, national, and international in scale. Statewide coverage was cited most commonly (43 occurrences), fol- lowed by corridor-specific and regional (26 and 25 citations respectively). One additional respondent specified corridor- specific across this respondent’s state. These tendencies gen- erally were prevalent across all survey types. Table 6 describes the modes that were covered in the surveys. Trucks dominated, at 105 responses, followed by rail (34), marine (28), intermodal (27), and air (24). Three respondents covered all modes. Again, these tendencies gen- erally were prevalent across all survey types. was air (93), and finally marine (90). Again, the domi- nance of trucking and the range of modes are consistent with the breadth of interest cited in the 2004 study (1). Table 2 does not include the following “other” • applications: Understanding the economic importance of trans- – portation facilities and needs Supporting the activities of a freight study or of a – freight task force (i.e., a freight council) Performance measures and management (from both – user and owner perspectives) Customer satisfaction and marketing among users – Provide “voice” to shippers – Informing plans and policies, including routing, – service coverage areas, bottlenecks, corridor stud- ies, master plans, systems plans, and goods move- ment action plans Assisting in data fusion and trend analysis (also a – function of the need to collect all types of data) Identifying gaps and priorities – Toll feasibility analysis – Reporting for the calculation of fees, including – statewide rail safety fee Analysis of the movement of hazardous materials – Administration of grants and loans – Calibration of a combined land use, economic and – transportation model Refinement of a strategic investment system plan. – The importance and broad application of data are encap-• sulated in one respondent’s comment that “the ways in which decision makers seek freight data and ask ques- tion is limitless. There are no areas of freight data that do not apply. The proprietary nature of freight data and the myriad of freight transportation projects [dictate] that freight data of all types [must] be captured.” Table 2 describes both data that were collected by the respondent and data that were assembled from other sources. In comparison, only 37 of the 56 respondents (66%) indi- cated that they actually administer or fund the data collec- tion: in other words, a significant number of respondents use data from other sources, and agencies may both collect and assemble data. There was a range of activities among the respondents who administered or funded surveys and data collection. The practitioners’ survey indicates that at least some agencies sup- ported more than one type of survey and data activity. As Figure 1 indicates, roadside/intercept surveys were most fre- quent (25 citations), followed by focus and stakeholder groups and personal interviews (18 citations each) and almost half conducted personal interviews (17 respondents). Next were mail-out/mail-back surveys (14), telephone surveys (12), and combined telephone and mail-out/mail-back surveys (7). Among newer electronic technologies, Internet (web) surveys were cited by 12 respondents, electronic license plate match-

14 T A B L E 3 S U R V E Y A P P L IC A T IO N S R oa ds id e/ in te rc ep t su rv ey s Te le ph on e M ai l- ou t/ m ai l- ba ck C om - bi ne d te le - ph on e/ m ai l- ba ck C om - m er ci al ve hi cl e tr ip d ia ry su rv ey s In te rn et su rv ey s Pe rs on al in te r- vi ew s Fo cu s/ S ta ke ho ld er gr ou ps G P S ve hi cl e tr ac ki ng su rv ey s L ic en se pl at e m at ch – m an ua l L ic en se pl at e m at ch – el ec - tr on ic A dm in i- st ra ti ve O th er To ta l L an d U se P la nn in g 3 4 1 1 0 1 7 6 1 0 2 0 0 26 In fr as tr uc tu re /F ac il it y P la nn in g 24 9 10 6 0 3 11 12 2 2 3 1 5 88 T ra ffi c S af et y O pe ra ti on s 13 5 4 3 0 3 5 7 2 2 2 1 2 49 D em an d M an ag em en t 17 5 5 2 2 1 7 7 2 0 2 1 1 52 A ir Q ua li ty M an ag em en t 6 3 2 1 1 0 3 4 1 0 0 1 2 24 L og is ti cs P la nn in g 8 3 4 3 0 1 6 6 0 0 1 1 1 34 M od el in g 1 1 2 V al id at io n of O th er D at a 3 1 1 5 R ou ti ng A na ly si s 2 2 E co no m ic I m pa ct /P er - fo rm an ce / A ct iv it y 2 2 F ac il it y M an ag em en t 1 1 1 3 C om m un ic at io ns 1 1 2 P ol ic y/ P la nn in g/ P ro gr am m in g 1 1 2 C us to m er S at is fa ct io n/ B us in es s D ev el op m en t/ M ar ke t R es ea rc h 1 1 1 1 1 5 O ut re ac h/ P ot en ti al P ro j- ec ts /P er ce pt io ns 1 1 2 4 O ri gi n- D es ti na ti on D at a fo r F re ig ht 1 1 1 3 V ar ie ty 1 1 2 O th er , N ot S pe ci fi ed 1 1 T ot al 79 34 29 20 4 11 43 43 9 4 10 5 15 30 6

15 T A B L E 4 IN C ID E N C E O F J O IN T S U R V E Y S R oa ds id e/ in te rc ep t su rv ey s Te le - ph on e M ai l- ou t/ m ai l- ba ck C om -b in ed te le ph on e/ m ai l- ba ck C om -m er ci al ve hi cl e tr ip di ar y su rv ey s In te rn et su rv ey s Pe rs on al in te r- vi ew s Fo cu s/ S ta ke ho ld er gr ou ps G P S ve hi cl e tr ac ki ng su rv ey s L ic en se pl at e m at ch – m an ua l L ic en se pl at e m at ch – el ec tr on ic A dm in i- st ra ti ve O th er To ta l R oa ds id e/ In te rc ep t S ur ve ys 7 0 1 0 0 0 3 3 1 3 1 1 0 20 T el ep ho ne S ur ve ys 3 5 3 2 1 0 6 2 1 1 0 0 2 26 M ai l- ou t/ M ai l- ba ck S ur ve ys 3 2 2 1 0 1 4 3 0 1 0 0 0 17 C om bi ne d T el ep ho ne M ai l- ou t/ M ai l- ba ck 2 1 2 1 0 0 1 1 0 0 1 0 0 9 C om m er ci al V eh ic le T ri p D ia ri es ( e. g. , t ri p lo gs ) 1 1 0 0 1 0 0 0 0 0 0 0 0 3 In te rn et S ur ve ys 0 1 0 1 0 0 1 1 1 0 0 0 0 5 P er so na l I nt er vi ew s 9 4 4 1 0 2 5 3 1 1 0 0 1 31 F oc us a nd S ta ke ho ld er G ro up s 3 3 4 0 0 2 4 3 1 1 2 0 2 25 G P S V eh ic le T ra ck in g 0 0 0 0 0 0 0 0 0 0 0 0 0 0 L ic en se P la te M at ch — M an ua l 1 1 0 0 0 0 0 0 0 0 0 0 0 2 L ic en se P la te M at ch — E le ct ro ni c 8 0 0 0 0 0 0 0 0 0 0 0 0 8 T ra ffi c C ou nt s 0 2 2 1 0 1 1 2 1 0 1 0 0 11 A dm in is tr at iv e D at a 3 0 0 0 0 0 1 0 0 0 0 0 0 4 O th er ( P IE R S d at a) 0 0 0 0 0 0 0 0 0 0 0 0 1 1 T ot al 40 20 18 7 2 6 26 18 6 7 5 1 6 16 2

16 T A B L E 6 M O D E S S U R V E Y E D R oa ds id e/ in te rc ep t su rv ey s Te le - ph on e M ai l- ou t/ m ai l- ba ck C om bi ne d te le ph on e/ m ai l- ba ck C om m er ci al ve hi cl e tr ip d ia ry su rv ey s In te rn et su rv ey s Pe rs on al in te rv ie w s Fo cu s/ st ak e- ho ld er g ro up s G P S v eh ic le tr ac ki ng su rv ey s L ic en se pl at e m at ch – m an ua l L ic en se p la te m at ch – e le ct ro ni c A dm in i- st ra ti ve O th er To ta l T ru ck 25 10 10 4 4 2 12 14 3 7 7 2 5 10 5 R ai l 4 5 1 2 7 11 0 1 3 34 A ir 4 4 0 2 7 6 0 1 0 24 M ar in e 5 5 1 2 8 5 0 1 1 28 In te rm od al 2 4 2 3 4 9 0 1 2 27 A ll M od es 1 0 0 2 3 S hi pp er s/ M ar ke t (a ll m od es ) 1 1 2 A ut os a nd B us es 1 1 O th er ( no t sp ec ifi ed ) 1 1 T ot al 25 26 29 9 4 11 39 45 3 7 7 6 14 22 5 T A B L E 5 G E O G R A P H IC S C O P E O F S U R V E Y S R oa ds id e/ in te rc ep t su rv ey s Te le ph on e M ai l- ou t/ m ai l- ba ck C om bi ne d te le ph on e/ m ai l- ba ck C om m er ci al ve hi cl e tr ip d ia ry su rv ey s In te rn et su rv ey s Pe rs on al in te rv ie w s Fo cu s/ st ak eh ol de r gr ou ps G P S v eh ic le tr ac ki ng su rv ey s L ic en se pl at e m at ch – m an ua l L ic en se pl at e m at ch – el ec tr on ic A dm in i- st ra ti ve O th er To ta l F ac il it y- sp ec ifi c (b ri dg es a nd tu nn el s) 1 0 0 0 0 0 0 0 0 0 0 0 1 2 F ac il it y- sp ec ifi c (m ar in e po rt s) 0 0 0 1 0 0 0 0 0 0 0 0 0 1 C or ri do r- sp ec ifi c 7 1 3 1 0 1 4 4 1 2 1 0 1 26 U rb an A re a 2 0 1 1 0 0 0 0 1 0 1 0 0 6 R eg io na l 5 5 4 0 2 0 2 3 0 1 1 0 2 25 S ta te w id e 10 5 4 4 0 3 6 8 0 0 0 2 1 43 S ta te w id e an d C or ri do r- sp ec ifi c 1 0 0 0 0 0 0 0 0 0 0 0 0 1 N at io na l 0 1 0 0 0 0 1 0 0 0 0 0 0 2 In te rn at io na l 1 0 0 0 0 0 1 0 1 0 0 0 0 3 T ot al 27 12 12 7 2 4 14 15 3 3 3 2 5 10 9

17 T A B L E 7 O R G A N Iz A T IO N S S U R V E Y E D R oa ds id e/ in te rc ep t su rv ey s Te le ph on e M ai l- ou t/ M ai l- ba ck C om bi ne d te le ph on e/ m ai l- ba ck C om m er ci al ve hi cl e tr ip di ar y su rv ey s In te rn et su rv ey s P er so na l in te rv ie w s Fo cu s/ S ta ke ho ld er gr ou ps G P S v eh ic le tr ac ki ng su rv ey s L ic en se pl at e m at ch – m an ua l L ic en se pl at e m at ch – el ec tr on ic A dm in is tr at iv e O th er To ta l V eh ic le O pe ra to rs 5 6 4 0 1 9 9 2 2 2 2 3 45 S hi pp er s/ R ec ei ve rs 8 7 3 1 2 11 11 0 1 1 1 3 49 3P L s 5 3 1 0 1 3 8 0 0 0 1 3 25 S er vi ce V eh ic le s 1 0 0 1 0 1 1 0 0 0 1 1 6 T er m in al s/ P or ts 6 5 1 0 2 7 10 1 1 0 1 3 37 D is tr ib ut io n C en te rs 4 3 1 1 1 5 9 0 0 0 1 3 28 O th er 0 0 0 O ce an C ar ri er s 1 1 P ub li c 1 1 2 M an uf ac tu ri ng , W ar eh ou se , R et ai l, T ra ns po rt at io n 2 1 1 4 C om m un it E co - no m ic D ev el op m en t 1 1 2 P la nn in g A ge nc y 2 2 P as se ng er V eh ic le s 1 1 R ai lr oa ds , M ot or - C ar ri er s, T ra il er 1 1 1 1 4 F re ig ht F or w ar d- er s, 4 P L s 1 1 F ac il it y O w ne rs 1 1 D ra ya ge O pe ra to rs 1 1 B or de r M an ag er s 1 1 T ra ns po rt at io n- M an ag er s 1 1 M il it ar y 1 1 C or ri do r A dv o- ca cy G ro up 1 1 U se rs 1 1 T ot al 0 33 27 12 4 9 38 54 3 4 4 7 19 21 4

18 the current trip: specifically, the trip origin and destination, goods carried (and their characteristics), and vehicle type (2). Some surveys also ask about the vehicle’s route. These surveys normally involve working with police or appropriate law enforcement agency to pull over moving vehicles/drivers and interview them at the roadside about their current trip. They also can be conducted at off-road locations such as weigh stations. The surveys usually are rel- atively brief, so as not to disrupt drivers and to avoid causing unnecessary traffic congestion. Roadside surveys, although most commonly cited by practitioners, are not used as often as in the past because of cost and the need for other agency involvement (e.g., law enforcement) (2). Attributes Twenty-three respondents provided details regarding their roadside/intercept surveys. Appendix A, a web-only docu- ment, contains the complete responses. Key points to note are as follows: Roadside/intercept surveys were used most commonly • for inter-urban goods movement [20 citations; 35 cita- tions if cross-border (9), international (6), and rural (2) were included]. Thirteen respondents indicated that these surveys were used for urban goods movement. A mix of information was collected, with some respon-• dents indicating that only origin-destination data were collected. Others also collected data on the goods and cargo, although to different extents. The data were used to corroborate truck counts: however, one respon- dent noted that they were “basically doing counts” and needed assistance from the state DOT to collect more data. Surveys were conducted at different locations, includ-• ing toll plazas, weigh stations, and at the entrance to a marine port (which served ocean-going vessels). One respondent noted that the roadside survey was “dif-• ficult and cumbersome, [but] very valuable.” Another respondent noted strong cooperation from drivers. The sample size varied, with the largest numbers of • respondents indicating samples of between 1,000 and 9,999 vehicles (eight responses), and between 100 and 999 (nine responses). (The question was posed to respondents in terms of categories of ranges.) One respondent in this group noted that this represented a 10% sample (part of a statewide study). One respon- dent indicated a sample size of between 10,000 and 49,000 vehicles, with the surveys conducted at toll plazas (i.e., when the vehicle was stopped); and this respondent indicated that a change to electronic toll collection technology would require a new method, including the use of mail-back surveys in addition to roadside interviews. No larger or smaller sample sizes Table 7 describes the types of individuals or organiza- tions that were surveyed. A broad range of interests is appar- ent. Among the 214 citations, shippers/receivers and vehicle operators dominated (49 and 45 citations respectively), fol- lowed by terminals/ports (37), distribution centers (28), and third- and fourth-party logistics providers (26). Six service operators were cited, as were four public planning and eco- nomic development agencies. STATE OF THE PRACTICE The literature identifies several different types of surveys and, as discussed here, there are many ways to categorize and define these surveys. However, the project scope identi- fied 10 specific survey types to be used for this Synthesis (see also chapter one, Subject). Each is described in the following subsections. However, to better reflect the actual practice (i.e., as manifested by the 12 types described in the practitioners’ survey) several mod- ifications were made to the aforementioned list: Two sets of survey types were combined (roadside • interview and vehicle intercept, and personal interview and focus and stakeholder groups). The discussion of license plate matches was subdivided • into manual and electronic techniques. The discussion of mail-back and telephone surveys • was brought together but also expanded to allow for telephone/mail-back combinations. Commercial vehicle trip diary surveys (i.e., driver sur-• veys) were added as a new category. The Panel’s original categories have been reordered to link the discussion of similar types together and generally to provide a logical flow from most common to least common, traditional to new technologies, quantitative to qualitative, and technical to administrative. All told, this analysis resulted in 12 separate survey types. Each type is discussed in turn here. Drawing from the practitioners’ survey, each discussion includes a defini- tion and a discussion of the key attributes (including “lessons learned,” frequency, and sample) with selected categories augmented by case studies where appropriate. A 13th cat- egory is provided for “other” survey types that were not oth- erwise included. Roadside/Intercept Surveys Survey Description Roadside/intercept surveys involve face-to-face interview surveys of vehicle drivers along a road or highway. These are typically used to capture data about the characteristics of

19 keep the surveys “as organized as possible when it comes to data entry and coding [as there is a] high potential for mistakes/human error.” Another respondent noted that “so much more” – needed to be done in terms of surveys, given that “basically” the only data collected were counts. Finally, one respondent acknowledged the complex- – ity of freight surveys, by noting that “the devil is in the details.” Respondents also noted several “critical” types of data • that were missing from the collected survey data: Off-hour/night-time trips—that is, the full 24-hour – data. In one case, “limited resources” were cited as the reason these data were not collected. Additional days [beyond the day(s) surveyed] – Small sample size (i.e., much activity is not – captured) Information on final destination, in cross-border – surveys Loaded versus empty movements – Detailed addresses for origin and destination infor- – mation. These were not collected because of “legal” considerations. Small trucks and intraregional trips, which were not – captured because the roadside surveys were con- ducted at weigh stations. Characteristics of the goods being transported – Movements at more “ideal” (i.e., representative) – times of year Rest area activity over a full 24-hour period – Logistics details. – Respondents expressed a wide range of comments on • the quality of the collected data. Some expressed sat- isfaction with that data: “we got what we were looking for;” “very useful and representative,” “statistically significant based on 8–9 daytime hours,” “helped to focus on specific areas of need.” Others were less satis- fied with the “usefulness, completeness [or] represen- tativeness” of the data: One respondent noted that the data represented only – a single day. Another commented that the data were incomplete, but these were “the best available.” A third noted that the quality was adequate but the sample size was too small “to do appropriate statis- tical analysis.” “Better survey mechanism[s]” was needed to – improve quality. One respondent expressed quality in varying – degrees: “about 90% [satisfaction] on origin and destinations; about 80–90% on routing; less than 60% on commodities.” One respondent noted that the “quality of the data – was useful and [thorough].” The data were captured using personal digital assistants (PDAs), and that each new set of data was downloaded regularly into a database and then reviewed for “potential errors.” were identified. Five respondents indicated they were unsure or did not respond. Each respondent had conducted roadside survey within • the past 9 years, including three conducted in the year of this study. The respondent with the oldest survey (2000) planned to conduct another one in 2009. However, most (10) surveys were conducted infre-• quently, on a one-time basis or only as needed. Seven respondents conducted roadside surveys every 5 years, and one conducted surveys every 10 years. Three respondents conducted roadside surveys annually. Five respondents indicated this was the first time the survey had been conducted. Respondents reported several successes with the • surveys: Some respondents indicated that they had collected – significant amounts of data. The surveys produced unanticipated findings. One – respondent gathered “new viewpoints on inter- modal facilities from shippers as well as [on] rail programs.” Another respondent “discovered truck route information from drivers that differed from planning assumptions.” Another respondent reported “average successes” – with the survey; however, “limited funding” restricted the range of studies to which the data could be applied. One respondent “successfully sampled approxi- – mately 10% of the daily traffic at all locations.” Another cited the importance of incentives in achieving a 15% to 20% response rate. Another success reported was using privately oper- – ated truck stops as data collection points (i.e., as opposed to public facilities such as weigh stations or by the side of the road). Respondents also reported some problems and • challenges: One respondent, although satisfied with the response – rate, noted that 10% of the data were “lost,” and that “interviewers’ knowledge of the system and the [trucking] industry, along with critical think- ing [and] reflective listening, [are] critical for valid data.” Another respondent noted the need to appro- priate personnel to conduct personal surveys. Another respondent noted the “huge” challenge in – obtaining vehicle registration data for use in deter- mining origin-destination patterns. Two respondents noted that the collected data were – limited to inter-urban flows (although these data were “rich”). Accordingly, intraregional and urban goods data were not captured, and corridor cover- age was “incomplete.” One respondent noted that the location of the sur- – veys also was important. One respondent noted that, although a “tremendous – amount of data” was gathered, there was a need to

20 reflected a “local context” that must be translated into a “statewide view.” The third respondent noted the chal- lenge of developing a survey “that is short enough to get [the] desired information without being too long to decrease participation.” Finally, respondents generally expressed satisfaction • with their surveys: “includes relevant issues” (i.e., no “critical” data were missing), “useful,” “[had] a good representative sample.” One respondent cited the mer- its of “effective tracking [methods]” to verify survey responses as an important factor. Telephone Interview Surveys Survey Description Telephone surveys gather information from a selected respondent through a telephone call in which the inter- viewer poses a series of pre-set questions and records the respondents’ answers. The respondent may or may not be notified in advance (e.g., by mail) and may or may not be pre-screened to ensure eligibility or to set up an appointment for the telephone call in order to allow the respondent time to prepare. The interviewer may record responses on paper for subsequent coding or directly into a computer, which in turn can validate responses in real time or prompt the inter- viewer to solicit corrections or probe further. Questions can be quantitative and/or qualitative. Attributes The practitioners’ survey yielded 12 responses to this ques- tion. Key points were as follows: The telephone surveys were conducted for a variety of • purposes, ranging from freight stakeholder outreach surveys that solicited input regarding potential proj- ects and motor carrier satisfaction surveys (i.e., with services provided by a government agency) to surveys conducted as part of MPO and statewide freight plans. One respondent noted that the survey had two main purposes: to gather “qualitative and quantitative infor- mation from stakeholders” and “to provide state and local transportation officials an opportunity to interact with freight stakeholders to learn more about freight considerations.” Although some surveys focused on the identification • of issues, others gathered quantitative and fact-based information about the respondent, such as (in this case, from a survey of firms’ business operations): location, industry type, facility type, number of trucks by size, truck type and ownership, and number of employees; the respondent’s trip-making characteristics also were solicited. Most surveys had been conducted within the past 3 • years, although two were conducted as early as 2003. Finally, one respondent noted the value of combining • survey results into a regional/national truck [origin- destination] database. However, “this would require shared standards on [method] and definitions.” Combined Telephone/Mail-Back Surveys Survey Description This method combines the use of telephone and mail-back surveys. Generally, selected respondents are called by tele- phone to screen their eligibility for the survey (for example, by ascertaining that they generate freight activity), verify the appropriate contact person to whom the survey should be directed, and solicit their participation. Qualified and participating respondents are then sent a paper survey form (or the form is delivered to them) to be returned by mail (or picked up). In some surveys, respondents have provided their information to the surveyor by telephone in lieu of the paper survey form. Attributes The practitioners’ survey yielded seven responses to this cat- egory. There was considerable variation in the attributes of the seven surveys. Key points to note are the following: Five of the respondents had conducted combined tele-• phone/mail-back surveys, most recently in 2009 (two respondents) and as far back as 2003. The seventh sur- vey had not yet been implemented. The survey frequency also varied, from annually or • biannually for the customer satisfaction surveys and up to every 5 years. Two respondents indicated they were unsure or that there was no established frequency. The survey purposes varied. Some surveys had more • than purpose and application. These included customer satisfaction surveys (two respondents), a statewide long range plan, a statewide transportation demand man- agement plan, facility management, economic impact assessment of road closures, and communications. The greatest number of these surveys (four) had state-• wide coverage. Corridor-specific, port-specific and urban each was cited once. The sample size and sampling frame also varied, • according to the purpose of the survey. One customer satisfaction survey sampled 100 to 999 firms; the other sampled 1,000 to 9,999 people. The statewide transpor- tation plan sampled 1,000 to 9,999 people. An ad hoc survey sampled one to 99 vehicles. Three respondents identified specific successes and • lessons learned for combined telephone/mail-back sur- veys. One respondent’s survey provided an “effective benchmark and tracking of customer service needs and delivery.” Another respondent commented that the sur- vey yielded “good information”; however, the results

21 The frequency ranged between once or twice yearly • to every 5 years. However, most respondents indicated that these surveys were conducted as needed, randomly or once only. The surveys were mainly regional or statewide in geo-• graphic scope, with five respondents each. One survey was focused on a specific corridor, and another was national in coverage. Most of the telephone surveys targeted business estab-• lishments (8). Five surveys had sample sizes of 100 to 999 firms and three had sample sizes of one to 99 firms. Several lessons were learned, notably:• One respondent noted that the telephone surveys – were time consuming, with some responses incom- plete. There also was difficulty in getting the appro- priate person to answer the questions. A second respondent attributed a relatively low response rate to respondents’ inability to answer key questions. A third respondent noted that response rates could be improved with a shorter survey, with interviews no more than 10 minutes long. Another respondent noted that stakeholders’ time – constraints often dictated their method of response. “The surveys were intended to be in person, but the stakeholders preferred conducting the interview over the phone. In some cases, the survey questions were answered via email interaction due to stake- holders’ time constraints.” A fifth respondent noted variation in the responses – to a statewide freight outreach survey owing to “public/political differences” across the state. Critical missing data including routing information • and “more survey respondents.” Respondents generally were satisfied with the quality • of their data, although with qualifications: Respondents’ willingness to divulge information – about themselves varied (i.e., business operations). Respondents’ ability to provide specific information – also varied. In one survey, whose sample was designed to ensure – that a wide range of shipper (industry) types was covered, a good “cross sample of data was obtained.” However, several major goods generators did not respond. Mail-out/Mail-back Surveys Survey Description Mail-out/mail-back surveys gather information by mailing a form for the respondent to complete and mail back. Often, a cover letter explaining the purpose of the survey and a post- age-paid return envelope are included with the form. These surveys are passive (i.e., self-administered), although assis- tance may be available by telephone or Internet and typically a set of instructions is included with the form. The advan- tage of this method to the survey sponsor is its relatively low unit cost compared with some methods (e.g., in-person or telephone calls); however, there is little ability to screen eli- gibility among respondents. This survey type also has been used in lieu of roadside surveys in locations where it would be unsafe to divert drivers (e.g., high-volume expressways) or where confidentiality concerns prohibited direct inter- views: this requires the recording of a license plate num- ber, the registered owner of which is then mailed a survey form. From the respondents’ perspective, this survey method offers some flexibility and ease in that it does not require an Internet connection and can easily be passed among differ- ent respondents within the same organization. On the other hand, erroneous responses can only be cleaned and corrected once the completed survey has been returned, and data must be coded manually or scanned into a computer. Questions can be quantitative and/or qualitative. Attributes The practitioners’ survey yielded 12 responses to this ques- tion. Key points were as follows: The mail-out/mail-back surveys were conducted for a • variety of purposes. These included freight stakeholder outreach surveys, consultation for statewide freight plans, roadside surveys (as described earlier), and in combination with telephone surveys “to increase response.” Surveys captured both quantitative data (freight • origin-destinations), whereas others gathered quali- tative information: “qualitative system evaluation,” input on “potential projects,” and “perceptions of transportation.” Most surveys had been conducted within the past 3 • years, although one was conducted as early as 2003. Almost all the mail-out/mail-back surveys were con-• ducted infrequently, as needed or once only. One respondent indicated that the survey was conducted every 5 years “at most.” The surveys were mainly regional or statewide in geo-• graphic scope, with four respondents each. Three focused on specific corridors and one covered an urban area. The mail-out/mail-back surveys targeted firms and • vehicles equally (four citations each), although freight stakeholders (“people”) also were sampled. Five sur- veys had sample sizes of 100 to 999 firms or vehicles, and four had sample sizes of one to 99 firms, vehicles, or people. A roadside mail-out/mail-back survey had a sample of 10,000 to 49,000 vehicles. One respondent noted that the surveys were successful • in gaining insight into what the “major shipper[s were] thinking.” However, others cited a mixed experience: One respondent cited a 10% return rate but noted – that some of the responses indicated that the

22 Almost all the personal interview surveys had been • conducted within the past 3 years, with the oldest sur- vey conducted in 2005. Survey frequency varied from continuous and “quar-• terly,” to an as-needed basis, and approximately every 5 years. The surveys were mainly statewide in geographic • scope (six citations) or corridor-specific (four cita- tions). There were two regional personal interview surveys and one occurrence each of national and inter- national personal interview surveys. Seven surveys had sample sizes of one to 99 individu-• als, and three each had sample sizes 100 to 999 and 1,000 to 9,999 (the latter comprising one continuous survey and one internal “as-needed” survey). Respondents noted several successes with the surveys. • One respondent noted that personal interviews were the “most important activity to determine priorities for policy and investment improvements.” Another respondent noted that, although the personal interview surveys were limited, “some information is better than none.” Lessons learned included the following: One respondent noted that the personal interviews – were “expensive and time consuming.” A second respondent noted the need for a “better mechanism” to gather this type of information. Another noted the difficulty of obtaining “in- – depth information from shippers, e.g., competitive issues.” A third respondent noted that its internal survey – provided “good background information and insight into complex issues,” although the respondent acknowledged that these [internal] surveys provided only “limited perspective of the situation.” Critical missing data comprised information that was • missing “due to proprietary data problems,” as well as a lack of specificity. Respondents generally were satisfied with the quality • of their data. One respondent noted that the information was used to support quantitative data. Another respon- dent noted that the survey “was useful in identifying relevant issues” but recognized that its small sample size was not representative. A third noted that assur- ances of confidentiality precluded any follow-up with respondents regarding the issues they raised. A fourth respondent noted that interviews with facility manag- ers proved very effective. Finally, a fifth respondent noted that personal interviews were “good,” provided the appropriate people were available to respond. Internet Surveys Survey Description Internet surveys are conducted via the World Wide Web. Sampled respondents can be recruited via telephone, mail, intended respondents were not made clear, either in the cover letter or because of stakeholders’ misunderstanding. Another respondent noted the need to follow up – with respondents in order to achieve a good rate of return. Critical missing data comprised responses from “some • of the larger freight generators in the state.” Respondents generally had mixed reactions regarding • the quality of their data: One respondent indicated satisfaction with the sur- – vey, which was intended to gather “qualitative feed- back on [the adequacy of and issues with] system operations [and on] business needs and issues that the [state] DOT should be aware of.” Another respondent indicated that the data were – good “but not complete or thorough.” Finally, a third respondent indicated that the data – were of “poor quality” and the surveys must be “redone.” Personal Interview Surveys Survey Description Personal interview surveys gather information from a selected respondent through a telephone call or face-to-face interview in which the interviewer poses a series of pre-set questions and records the respondents’ answers. The respon- dent may or may not be notified in advance (e.g., by mail, telephone, or by being intercepted), and may or may not be pre-screened to ensure eligibility or to set up an appoint- ment for the interview in order to allow the respondent time to prepare. The interviewer may record responses on paper for subsequent coding or directly into a computer, which in turn can validate responses in real time or prompt the inter- viewer to solicit corrections or probe further. Questions can be quantitative and/or qualitative. Attributes The practitioners’ survey yielded 13 responses to this ques- tion. Key points were as follows: Personal interview surveys were conducted for a vari-• ety of purposes, ranging from continuous surveys and the gathering of information for specific initiatives or for updates of existing information (e.g., port and air- port volumes), to the development of a business plan for the distribution industry and a statewide freight plan. The surveys gathered qualitative input, which was used • for long-range planning and policies, assess customer satisfaction, promote business development, profile areas of economic activity, get a “pulse” of the system, and estimate future needs.

23 Focus and Stakeholder Group Surveys Survey Description Focus and stakeholder group surveys are used to solicit qualitative comments and perceptions regarding issues, new programs and similar. Focus group surveys often are con- ducted face-to-face among small groups. Stakeholder sur- veys can be conducted face-to-face or by telephone, mail, or Internet. Attributes Fifteen respondents indicated they had conducted focus and stakeholder group surveys. Key points were as follows: Focus and stakeholder group surveys were used for a • variety of purposes: input to freight plans (including study technical advisory committees), freight advisory committees or councils, issue-specific stakeholder consultation (e.g., for emissions and for a statewide truck-only lane initiative), outreach and communica- tions, and highway corridor studies. Twelve of the surveys had been conducted within the • past 3 years, with others conducted as early as 2002. The frequency of the surveys ranged from bimonthly • to 5 years, with a large number conducted as needed. The surveys were mainly statewide in geographic • scope (eight citations), with four corridor-specific sur- veys and three regional surveys. Most surveys were conducted among groups of one to 50 • participants (10 citations). Three were conducted with groups between 50 and 99 individuals, and two more were conducted with groups up to 499 participants. Among all the surveys described in this Synthesis, these had the most varied audiences: shippers/receiv- ers (11 citations), terminals/ports (10), vehicle opera- tors (9), distribution centers (9), third-party logistics providers (8), border managers, transportation manag- ers, service vehicles, rail and motor carriers, facility users, the military, and a corridor advocacy group. Respondents reported both successes and lessons • learned. Successes included: Ability to interface with stakeholders, both in – groups and via one-on-one interviews. Ability to bring together “business representatives – in the state” for a statewide freight planning study, from whom the respondent “learned quite a bit about planning issues that [the agency was] not consider- ing, particularly to plan beyond state borders.” One state DOT had “an active and ongoing participa- – tion in the planning processes of all federally-man- dated Metropolitan Planning Organizations in [the state], which function as the coordinating commit- tee for transportation and freight issues that bring together city and county governments, State DOTs or e-mail, then—if they meet eligibility criteria—they are directed to an Internet link to complete the survey. Par- ticipants also can be invited to participate directly, such as through a targeted e-mail list that points to a link (as in the case of the practitioners’ survey that was developed for this Synthesis). Finally, participation can be opened (uncon- trolled) with access provided to any respondent who wishes to participate, for example through an advertisement in user group publications. As with mail-out/mail-back surveys, this method also is passive, although its ability to reach a very large number of users can make its unit costs quite low. Internet surveys have been used to solicit qualitative input, and more recently have been used for quantitative surveys. Attributes Four respondents indicated they had conducted Internet sur- veys. Key points were as follows: Internet surveys were used mainly for freight/inter-• modal planning studies. One respondent noted that the survey was used to a • planning education and information dissemination program. Another respondent indicated that the survey was not used to gather freight data; the focus was on qualitative information. A third respondent noted that its Internet survey was used to gather quarterly data, yielding “valuable local and regional trends.” Three of the Internet surveys had been conducted • within the past 2 years, with the fourth conducted in 2005. Two of the surveys were conducted on an as-needed • basis, a third was conducted every 1 to 2 years, and the fourth survey was continuous. The surveys were mainly statewide in geographic scope • (three citations), with one corridor-specific survey. Each of the surveys had samples of different size, rang-• ing from one to 99 firms to 100,000 or more individu- als (this last for the continuous survey). Respondents noted several successes with the surveys. • Respondents use terms such as “valid” and “valuable” to describe the input. One respondent noted that “peo- ple prefer to take surveys via Internet” and that it was “easy to tabulate the results.” One lesson learned was that “some of the recipients do • not have access to a computer.” Critical missing data comprised the “volume of com-• modities shipped in pounds.” Respondents generally were satisfied with the qual-• ity of their data (“excellent,” “gives a more accurate picture of each corridor individually”). However, one respondent noted that although the data were useful, they were “not always complete.”

24 Commercial Vehicle Trip Diary Surveys Survey Description Commercial vehicle trip diary surveys are used to collect detailed information about the activities of a single vehicle, usually over a single day or a few days. They can provide data about exact locations served, route, arrival and depar- ture times, time taken for delivery/collection/servicing, type of goods/services, and the like. They typically are self- completed by the driver or by another suitably informed employee of the freight operator. These surveys gather infor- mation the characteristics of the trip (e.g., location of stops, activity at stops, arrival/departure times, itinerary, parking, goods transported). The driver or another vehicle occupant must record the activity at each stop (2). Attributes Two respondents indicated they had conducted commercial vehicle trip diary surveys. Key points were as follows: One trip diary survey was conducted for an MPO’s • truck travel and congestion study, and the other was conducted as part of another MPO’s urban goods movement data collection study as input to a truck trip model update. Both surveys were conducted in the past 2 years.• The truck travel and congestion study was conducted once • only; the trip diary survey was conducted every 4 years. Both surveys were regional in geographic scope.• The truck travel and congestion survey sampled 100 • to 999 vehicles; the trip diary survey sampled 1,000 to 9,999 vehicles. The truck travel and congestion survey was successful: • “The survey was designed to gather information [such as] location of facility; type of facility; industry type; number of trucks by size; type of trucks; truck ownership; number of employees at facility; percent of trips delivering to mul- tiple locations; percent of trips delivering to single loca- tion; [and] nature/land use of destinations. The survey duration was not more than 15 minutes, with a reasonably good response rate. These types of surveys have [proven] to be very effective for sectors like manufacturing, whole- sale, and warehousing, where trip making characteristics seem to involve a finite set [and number] of destinations or land use types, and also where the starting and ending point of trips seem to be at these facilities.” The other respondent noted, as a lesson learned, that the • “useful” number of responses was “very low,” with “81 of the 392 selected companies [returning] 362 diaries.” However, the collected data were “completely relevant • to [the] truck trip model update.” and representatives of U.S.DOT (Federal Highway and Transit Administrations). It is worth specifi- cally noting that [the State DOT] has full and active participation in [its] Freight Mobility Plan Technical Advisory Committee for the [state’s largest MPO] and the Metropolitan Planning Organizations for [some marine port cities]. The [State DOT] also [has a] close working relationship with the [state ports authority, which manages the aforementioned ports], which functions as the liaison with shippers, terminal operators, etc.” One respondent reported that the results of the – stakeholder “forum” became the starting point for a statewide freight advisory committee, which in turn supported the statewide freight plan. Several lessons also were reported:• There was a need to better manage stakeholders’ – expectations. Some respondents noted that the focus and stake- – holder group surveys had limited success, for sev- eral reasons: difficulty in getting representatives of the invited stakeholders to attend the meetings, lack of follow-through, politically charged meet- ings; “modest penetration” of the stakeholder group, and meetings that were dominated by “major players.” One respondent noted that one of the stakeholders (the state motor transport asso- ciation) was a “voting member” of the initiative’s steering committee. Critical missing data were identified as more infor- – mation regarding the supply chain (final destina- tion of the goods being transported), the logistical needs of the private sector, and other nontrucking modes (i.e., rail). One respondent commented that its focus and stakeholder group surveys should have been supported by other efforts to get more participation. Respondents expressed mixed satisfaction with the – quality of their surveys. One respondent found the results “modestly useful,” and another noted that the low turnout and lack of input from the attend- ees limited the usability of the results. A third noted the need for the survey to be part of a larger effort. However, one respondent noted that the information was “specifically” useful in working with the pri- vate sector and in reconciling the varied planning horizons of interest among the different partici- pants. Another noted that the concerns of stake- holders were helpful to the planning process. A third respondent noted the “excellent quality [of the surveys, which yielded] great first-hand accounts and a good network to follow the logistics paths in the state.”

25 is traced—for example, along a section of an expressway, whereby the entrance and exit interchanges can be identi- fied along with through traffic. Through expansion of the surveys with traffic counts, an origin-destination matrix for the facility or corridor can be identified. Manual license plate matches involve the recording of the license plate data and, as appropriate, the vehicle type, by the human eye. The data can be recorded on paper or electronically (e.g., into a spreadsheet), or through audio recordings. License plate surveys have the advantage of being unobtrusive to the trav- eling public. Attributes Three respondents reported their manual license plate match surveys. Key points were as follows: Surveys were both old (2003) and recent (2009).• The surveys were conducted on an as-needed basis.• Two surveys were corridor-specific, and a third cov-• ered a region. One survey recorded a sample size of one to 99 vehi-• cles, a second survey had a sample size of 100 to 999 vehicles, and a third survey recorded a sample size of 1,000 to 9,999 vehicles. License Plate Match Surveys—Electronic Survey Description These surveys are similar to manual license plate match surveys, except that the data are recorded electronically using video recorders. The recorded data subsequently are transcribed manually into spreadsheets for processing. The use of recording devices can allow for greater accuracy in the records, reduced labor costs, and larger sample sizes. It also provides a permanent record of the observations. Newer technologies digitize and translate the data directly, thereby eliminating the manual transcribing step. Attributes Three respondents reported their electronic license plate match surveys. Key points were as follows: Surveys were conducted at different times, including • as far back as 2003. Most surveys were conducted as needed, although one • was conducted every 5 years. One survey was corridor-specific, a second covered an • urban area, and a third covered a region. Two surveys recorded sample sizes of 1,000 to 9,999 • vehicles. A third survey had a sample size of between 50,000 and 99,999 vehicles. One respondent noted the benefits as providing “an • excellent source for identifying a corridor traffic pro- One respondent commented that response rates could • be improved by having a shorter survey, with less than 10 minutes per interview. Global Positioning System Vehicle Tracking Surveys Survey Description GPS refers to “a federal system of satellites which allow the user to pinpoint any location using triangulation” (3). The latitude and longitude of a GPS receiver (which may be located, for example, onboard a vehicle or in a cell phone) can be determined through satellite transmissions. Attributes Five respondents indicated they had conducted GPS surveys. Key points were as follows: One survey measured truck performance. Two other • GPS surveys were conducted to track truck border crossing times, with one of these surveys also record- ing routes. The border crossing time/routing survey is ongoing • and continuous. The other border crossing study was conducted in 2009. The truck performance study was underway at the time of the response. The truck performance study covered an urban area; • the two border crossing studies were described as cov- ering a specific corridor and an international area. The other border crossing study and the truck perfor-• mance study each recorded samples of 100 to 999 vehi- cles. No sample size was given for the third survey. The border crossing time/routing survey was “very • effective.” However, the other border crossing study of wait times only provided “adequate to good” data. For the truck performance measurement study—which • was a pilot effort—the respondent noted that it was “challenging and time consuming to secure access to the [GPS providers’] data and [to get the appropriate] contracts in place.” However, the respondent also noted that “companies have been very willing to share data and participate” in the study. Respondents found the quality of the collected data to • be “reasonable” and “very good,” although one respon- dent noted that information differentiating travel times by time of day was missing. License Plate Match Surveys—Manual Survey Description License plate match surveys involve the recording of all or part of a vehicle’s license plate as it passes through two or more points along one or many facilities or corridors. In this way, the vehicle’s movement through the facility or corridor

26 file.” A second respondent noted that these surveys “[do] not interrupt the traveling public.” Respondents also noted some limitations to the data: • Two respondents noted that the data were samples only, with one respondent observing that “plate surveys are taken only 2–3 times in a week [so] data [are] often not [a] representative sample of the traveling public.” Another respondent further noted that the data were “good for a marketing survey, but due to time con- straints and budget considerations, [the survey] does not provide ... a good representative sample.” Finally, one respondent noted that these surveys “cannot com- pletely verify the actual origin and destination of the trip from which [the data were] captured [and] in terms of determining the % of through traffic [the survey data were] not very accurate.” Administrative Surveys Survey Description Administrative surveys refer to information-gathering exercises that capture data that are not related specifically to transportation planning—for example, vehicle owner- ship and registration, insurance, shipment value, and the like. These data may be collected for financial record- keeping, legal, security, insurance or other administrative purposes. Attributes Two respondents indicated they had conducted administra- tive surveys. Key points were as follows: The administrative surveys were conducted for a study • of drayage activity at a large port and for a freight mobility study. Both surveys were conducted in the past 2 years.• The port drayage survey was conducted once only; • the freight mobility survey was conducted every few years. Both surveys were statewide in geographic scope.• One survey sampled one to 99 individuals (who were • asked to report on tons and value of the goods being shipped); the other sampled 1,000 to 9,999 vehicles. No successes were reported. However, one respondent • noted that the completeness, representation, and use- fulness of the data were all satisfactory. One lesson learned was that “[the contracting pro-• cedure with an external] administrative department is very complicated and time consuming” (i.e., these arrangements were required in order to gather the req- uisite information). Critical missing data were commodity flows and data • detailed at the county level. Other Surveys Five respondents indicated that they conducted “other” types of surveys. These included: A truck toll revenue data survey, which recorded “basic • information” about the types of vehicles (i.e., number of axles) that used the respondent’s bridges and tun- nels by time of day and by method of payment. The survey is conducted continuously, and sampled more than 100,000 firms. A trucking company marketing survey, which • accounted for 10,000 to 49,999 vehicles. This survey is conducted as needed, and was last conducted in 2002. A product survey (in this case, a “virtual container • yard”), which sampled one to 99 firms in 2009 and is ongoing. The survey comprised “personal meetings and phone interviews with key stakeholders on prod- uct need, design, use and pricing is an iterative survey process as product is developed and tested.” SURVEY COSTS Respondents to the practitioners’ survey were asked to indi- cate the approximate cost range of their “last” survey. Table 8 summarizes the results. By far the most common cost range was less than $0.5 million (76 of 103 responses, or 74%). This was followed by $0.5 to $1.0 million (15 responses), and the remaining three indicated survey costs were for the $1.0 to $5.0 million range. Table 9 summarizes the allocation of the costs between internal resources and external resources (e.g., consultants or equipment). The allocations ranged between 0% (i.e., all costs were external) and 100% (there were no external costs), with a 20% internal split being the most common category (39 of 99 responses, or 39%). The responses sug- gest that most of the costs of data collection are external to the agency, although the 100% internal category represented almost one-quarter (23%) of the respondents. These findings should be considered with caution for sev- eral reasons: The responses to both questions were approximations • only. The responses represent categories, within which there • can be considerable variation. There can be variation in costs within a survey cat-• egory according to the exact nature of the activity. Finally, as one respondent commented, the data and • survey activities were conducted as part of a compre- hensive planning study, and it was difficult to separate the proportion of costs attributable to surveys and data collection from the overall study costs.

27 T A B L E 8 A P P R O X IM A T E C O S T O F L A S T S U R V E Y R oa ds id e/ in te rc ep t su rv ey s Te le ph on e M ai l- ou t/ m ai l- ba ck C om bi ne d te le ph on e/ m ai l- ba ck C om m er ci al ve hi cl e tr ip d ia ry su rv ey s In te rn et su rv ey s Pe rs on al in te rv ie w s Fo cu s/ st ak eh ol de r gr ou ps G P S v eh ic le tr ac ki ng su rv ey s L ic en se pl at e m at ch – m an ua l L ic en se pl at e m at ch – el ec tr on ic A dm in is tr at iv e O th er To ta l < $0 .5 m 16 9 8 5 1 4 11 11 1 3 2 2 3 76 $0 .5 –$ 1. 0 m 5 0 0 1 1 0 1 3 2 0 1 0 1 15 $1 .0 –$ 5. 0 m 1 1 1 0 0 0 0 0 0 0 0 0 0 3 $5 .0 –$ 10 .0 m 0 0 0 0 0 0 0 0 0 0 0 0 0 0 > $1 0. 0 m 0 0 0 0 0 0 0 0 0 0 0 0 0 0 N /A 1 1 4 0 0 0 1 0 0 0 0 0 2 9 T ot al 23 11 13 6 2 4 13 14 3 3 3 2 6 10 3 N /A = n ot a va il ab le . T A B L E 9 P E R C E N T A G E O F S U R V E Y C O S T S T H A T A R E I N T E R N A L R oa ds id e/ in te rc ep t su rv ey s Te le ph on e M ai l- ou t/ m ai l- ba ck C om bi ne d te le ph on e/ m ai l- ba ck C om m er ci al ve hi cl e tr ip d ia ry su rv ey s In te rn et su rv ey s Pe rs on al in te rv ie w s Fo cu s/ st ak eh ol de rs gr ou ps G P S ve hi cl e tr ac ki ng su rv ey s L ic en se pl at e m at ch – m an ua l L ic en se pl at e m at ch – el ec tr on ic A dm in i- st ra ti ve O th er To ta l 0% in te rn al 2 0 1 0 0 0 2 2 1 0 0 0 1 9 20 % in te rn al 9 4 3 5 1 2 3 7 0 0 1 1 3 39 40 % in te rn al 3 0 0 0 0 0 1 0 0 0 0 0 0 4 60 % in te rn al 3 2 1 0 0 0 1 1 0 0 0 0 0 8 80 % in te rn al 0 0 0 0 0 0 1 1 1 0 1 0 0 4 10 0% in te rn al 3 3 4 1 0 2 3 3 0 1 1 1 1 23 N /A 4 2 1 0 1 0 2 1 0 1 0 0 0 12 T ot al 24 11 10 6 2 4 13 15 2 2 3 2 5 99

28 inclusion of these other modes—rail, air, water, intermodal, and pipeline—provides both a context for this focus (i.e., in identifying respondents’ modal interests) and completeness (i.e., because the other modes may influence or be influenced by trucking). Table 13 indicates that although highway/truck planning dominates in respondents’ planning activities (36 of 39 responses), other modes—rail (29), marine (22), and air (21)—also figure in planning activities. Intermodal and multimodal freight also was important, at 25 responses. Pipeline freight was cited once. The next four tables describe the data needs for each mode in turn. Table 14 describes highway/truck data needs. Count data were used most commonly, at 33 citations, followed by vehicle size (26) and vehicle type (23). Other characteristics about the vehicle, its cargo and the trip also were used. Costs were not commonly used, although the need for these data was cited, with freight rate data at 17 citations and line-haul costs at 16 citations. More important among the needs were travel time reliability (21 citations), truck origin/destination patterns and number of stops (19 each), travel time (18), and trip origin/ destination patterns (17). Almost the same numbers of respon- dents who reported the need for a type of data did not need the data—notably, vehicle emission data and cost data. Table 15 summarizes the requirements for rail freight data. Origin/destination patterns and commodity data were used most commonly, at 20 citations each. For the remainder of the data types, the needs generally and often significantly exceeded the uses: stop/delay data (17 responses), routing data, travel time, and reliability (16 each), and ramp-to-ramp costs and freight rate (15 each) were the most common needs. Table 16 lists the freight data requirements for air. Com- modity and shipment data were most commonly used (11 responses each), followed by origin/destination patterns (seven) and routing data (five). Origin-destination patterns (13 responses) and routing, travel time, and reliability data (12 responses each) were the most commonly cited needs. Table 17 lists the data requirements for marine freight. Commodity data were used most frequently, at 19 citations, followed by origin/destination patterns (13), shipment data (12), and equipment data (10). The most common data needs were for reliability and port-to-port costs (10 citations each), followed by travel time (9). Finally, respondents were asked separately about inter- modal data uses and needs. Table 18 indicates that, by far, intermodal data for combinations of trucks and other modes were used or needed, with the truck/rail combination elic- iting 29 responses, followed by truck/marine (24) and truck/airport (21). The rail/marine combination elicited 21 responses. DATA AVAILABILITY AND DISSEMINATION Respondents to the practitioners’ survey were asked to describe the availability and dissemination of their survey results to the public and to other external agencies. These questions provide some indication of the ability of data ‘owners’ to share data and results. Table 10 indicates that the large majority of respondents (78 of 98 respondents, or 80%) made their data available to the public or to other external agencies. All but one of these respondents provided the data free of charge. Table 11 summarizes how the data were disseminated. Hardcopy reports and electronic data dissemination were both widely used, with the latter cited slightly more fre- quently (73 versus 69 citations respectively). (Note that respondents could respond to all applicable methods, mean- ing that many respondents made both hardcopy and elec- tronic reports available.) One respondent indicated that its data were available online. FREIGHT DATA REQUIREMENTS Section 3 of the survey asked respondents to describe the type of data they use or need. Table 12 lists seven different types of data, and distinguishes them according to whether or not respondents currently use the data, they need the data but the data are not available, or the data are neither used nor needed. The large majority of the 39 respondents indicated a current use or a need for each of the seven data types. Other key points to note: Commodity data (30 respondents) and origin/destina-• tion data (23 respondents) were used most commonly. Fewer than half the respondents used the other types of data. The lowest use was reported for data on terminal • and intermodal transfer facilities (six respondents). However, this category also represented the greatest need, at 21 respondents. Data on shipments (18 respondents) and routing (16 • respondents) were next most commonly cited as needs. The fewest needs were recorded for commodity data (three respondents). Cross-border data (nine respondents) and cargo data • (six respondents) were the data types most commonly cited as not being applicable. Origin/destination data (one respondent) and commodity data (two respon- dents) were least commonly cited, corresponding to common usage and need. Section 3 asked about data for modes other than trucks. Although the focus of this Synthesis is on truck surveys, the

29 T A B L E 1 1 D A T A D IS S E M IN A T IO N R oa ds id e/ in te rc ep t su rv ey s Te le ph on e M ai l- ou t/ M ai l- ba ck C om bi ne d te le ph on e/ m ai l- ba ck C om m er ci al ve hi cl e tr ip di ar y su rv ey s In te rn et su rv ey s Pe rs on al in te rv ie w s Fo cu s/ st ak eh ol de r gr ou ps G P S ve hi cl e tr ac ki ng su rv ey s L ic en se pl at e m at ch – m an ua l L ic en se pl at e m at ch – el ec tr on ic A dm in i- st ra ti ve O th er To ta l H ar dc op y (m ay al so in cl ud e pr es en ta ti on ) 14 7 8 4 1 3 11 10 2 1 3 2 3 69 E le ct ro ni c D at a F il e 17 9 6 4 2 3 10 9 3 1 3 2 4 73 S um m ar y O nl y 3 0 0 1 0 0 0 0 0 0 0 0 0 4 D is cu ss io n 0 2 1 0 0 0 1 0 0 0 0 0 0 4 O nl in e R ep or t/ W eb si te 0 0 0 0 0 0 0 1 0 0 0 0 0 1 S tu dy N ot Y et C om pl et e 0 1 0 0 1 0 0 1 0 0 0 0 1 4 O th er ( no t o th er - w is e id en ti fi ed ) 0 0 0 0 0 0 0 1 0 0 0 0 0 1 T ot al 34 19 15 9 4 6 22 22 5 2 6 4 8 15 6 T A B L E 1 0 D A T A A V A IL A B IL IT Y T O T H E P U B L IC O R T O E X T E R N A L A G E N C IE S R oa ds id e/ in te rc ep t su rv ey s Te le ph on e M ai l- ou t/ m ai l- ba ck C om bi ne d te le ph on e/ m ai l- ba ck C om m er ci al ve hi cl e tr ip di ar y su rv ey s In te rn et su rv ey s Pe rs on al in te rv ie w s Fo cu s/ st ak eh ol de r gr ou ps G P S ve hi cl e tr ac ki ng su rv ey s L ic en se pl at e m at ch – m an ua l L ic en se pl at e m at ch – el ec tr on ic A dm in i- st ra ti ve O th er To ta l Y es — N o C ha rg e 18 9 8 5 1 4 9 11 3 1 3 2 3 77 Y es — A t a P ri ce 0 0 0 1 0 0 0 0 0 0 0 0 0 1 N ot A va il ab le 5 2 3 0 1 0 4 3 0 1 0 0 1 20 T ot al 23 11 11 6 2 4 13 14 3 2 3 2 4 98

30 equivalent units). Finally, one respondent noted that much of the data must come from the private sector, which would incur additional cost. USE OF EXISTING DATA SETS Section 4 of the survey asked respondents to review a list of 21 public and commercial (private) American data sets that cov- ered the major freight modes (truck, rail, marine, and air). The purpose was to determine which of these existing sources, external to the respondents’ organizations, were used; that is, as a possible alternative to in-house data collection. Respon- dents also were asked to assess the data sets that they used. Thirty-five respondents indicated that they used these external data sets to “populate [their] freight databases.” Table 19 shows the usage of the 21 public and commercial data sources. The most frequently used data set was the Freight Analysis Framework (U.S.DOT, 33 responses), followed by the U.S. Commodity Flow Survey (31) and TRANSEARCH Insight Database (25). Table 20 describes the purpose for using the public and commercial data sets. The most common purpose was for infrastructure/facility planning (30 responses), followed by demand management (14), traffic safety and operations (10), and land use planning and logistics planning (9 each). There were also several “other” purposes, which comprised policy development, long range planning, corridor planning, proj- ect planning, modeling (including model updates), market research, cost benefit analysis, commodity flow analysis, freight statistics and studies. TABLE 12 FREIGHT DATA USED OR NEEDED Freight Data Currently Use Need But Not Available N/A (do not use or need) Commodity detail (e.g., formal classification system) 30 3 2 Cargo detail (e.g., aggregate categories, hazardous and nonhazardous cargo, empty vs. non-empty) 18 10 6 Origin-destination detail (e.g., states, zIP codes, counties, shipper detail, traffic analysis zone, customs port of exit/entry) 23 10 1 Shipment detail (e.g., weight, volume, value, mode of transport, average length of haul, number of stops per trip, time-sensitive shipment, truckload or less-than-truckload ship- ments, empty shipments) 13 18 4 Routing details (e.g., major routes used, number of stops, interim trip origin and destina- tions, vehicle routing, hazardous materials vehicle routing) 13 16 5 Cross border data (e.g., origin/destination patterns, commodity, vehicle type, shipment characteristics, mode, stop/delay data) 15 10 9 Terminal and intermodal transfer facilities (e.g., truck volumes entering/exiting, conges- tion-related delays on access roads, length of queue on access roads, incident rates on access roads, travel time contours around the facility, capacity of facility) 6 21 5 TABLE 13 MODES CONSIDERED IN PLANNING Mode Responses Truck/Highways 36 Rail 29 Air 21 Water (marine port, barge, short sea ship- ping, ferry) 22 Intermodal/Multimodal 25 Other (pipeline) 1 Respondents described their needs for intermodal data. Although some had data that covered certain (though not all) characteristics of each mode, such as origin, destination, vol- ume, and counts, the full multimodal profile and the details of the intermodal transfer were lacking. One respondent expressed these gaps as “upstream origin, downstream des- tination, transfer time, transfer cost, commodity, shipment details, intermodal facility location, and operation time. [Origin-destination information] for intermodal freight is often available for individual links, but not for the complete intermodal trip.” Other specific needs were intermodal move counts; number of container/trailer lifts; commodity ton- nage (weight), type, and value; routing (for all modes); and characteristics of dray trips, including local activity. In addition, respondents wanted activity measured for and differentiated by facility and non-facility cargo and vehicles. They wanted up-to-date data that could be available in dif- ferent measures, such as short tons and TEUs (twenty-foot

31 TABLE 16 AIR FREIGHT DATA USE AND NEEDS Data Type Currently Use Need But Not Available N/A O/D Patterns 7 13 1 Commodity 11 11 0 Shipment (weight, vol- ume, value) 11 8 1 Routing Data 5 12 4 Travel Time 2 12 6 Reliability 2 12 6 Air Freightage 1 11 8 Drayage Costs 2 11 7 Freight rate (e.g., cost per ton-mile) 1 11 8 Other 1 6 8 N/A = not available, O/D = origin/destination. TABLE 14 HIGHWAY/TRUCK FREIGHT DATA USE AND NEEDS Data Type Currently Use Need But Not Available N/A Vehicle Type 23 6 7 Vehicle Size 26 5 5 Average Vehicle Speed 16 9 11 Vehicle Emission Data 6 15 14 Traffic Counts and Classification Data 33 1 2 Cargo Type 19 10 5 Payload Weight 15 11 8 Shipment Value 12 13 8 Truck O/D Patterns 15 19 1 Trip O/D Patterns 16 17 1 Travel Time 11 18 6 Travel Time Reliability 5 21 10 Number of Truck Stops for LTL Shipments 0 19 15 Incident Data 17 7 11 Line-haul Costs 3 16 15 Drayage Costs 4 14 16 Freight Rate (e.g., cost per ton-mile) 4 17 13 Other 2 1 12 N/A = not available, O/D = origin/destination. TABLE 15 RAIL FREIGHT DATA USE AND NEEDS Data Type Currently Use Need But Not Available N/A O/D Patterns 20 9 0 Commodity 20 9 0 Equipment Details (e.g., car type) 12 11 6 Shipment (e.g., weight, volume, value) 11 12 6 Routing Data 9 16 4 Travel Time 4 16 9 Reliability 2 16 10 Stop/Delay Data 2 17 10 Ramp-to-ramp Costs 1 15 13 Freight Rate (e.g., cost per ton-mile) 4 15 10 Other 0 3 10 N/A = not available, O/D = origin/destination.

32 TABLE 19 USAGE OF PUBLIC/COMMERCIAL DATA SOURCES Public/Commercial Data Sources Do You Use? Yes Airport Activity Statistics of Certified Route Air Carriers—U.S. Bureau of Transportation Statistics 10 Border Crossing Data—U.S. Bureau of Trans- portation Statistics 18 Commodity Flow Survey (CFS)—U.S. Bureau of Transportation Statistics and the Census Bureau 31 Freight Analysis Framework (FAF)—U.S.DOT 33 Freight Commodity Statistics—Association of American Railroads 13 IANA Report—Intermodal Association of North America 3 Industry Trade Data and Analysis—Interna- tional Trade Administration, U.S. Department of Commerce 8 LTL Commodity and Market Flow Database— American Trucking Association 5 MARAD—U.S.DOT Maritime Administration 7 Maritime Administration Office of Statistical and Economic Analysis 7 National Roadside Survey/Commercial Vehi- cle Surveys 4 North American Trucking Survey (NATS)— Association of American Railroads 5 Port/Import/Export Reporting Service (PIERS)—Journal of Commerce 7 Rail Waybill Sample—Surface Transportation Board 22 RAILINC—American Association of Railroads 1 State Estimate of Truck Traffic—FHWA 9 Transborder Surface Freight Data—U.S. Bureau of Transportation Statistics 10 TRANSEARCH Insight Database 25 TransStats: The Intermodal Transportation Database—U.S. Bureau of Transportation Statistics 8 Vehicle Inventory and Use Survey (VIUS)— U.S. Census Bureau (discontinued as of 2002) 8 Waterborne Commerce of the United States— U.S. Army Corps of Engineers 21 USE OF INTELLIGENT TRANSPORTATION SYSTEM TECHNOLOGIES The potential of ITS (Intelligent Transportation System) technologies to reduce data collection costs, increase sample TABLE 17 MARINE FREIGHT DATA USE AND NEEDS Data Type Currently Use Need But Not Available N/A O/D Patterns 13 5 4 Commodity 19 3 0 Equipment Details (e.g., vessel type) 10 4 7 Shipment (e.g., weight, volume, value) 12 6 3 Routing Data 7 7 7 Travel Time 5 9 8 Reliability 4 10 8 Port-to-port Costs 4 10 8 Drayage Costs 6 7 8 Freight Rate (e.g., cost per ton- mile) 6 8 7 Other 7 4 8 N/A = not available, O/D = origin/destination. TABLE 18 INTERMODAL FREIGHT DATA USES AND NEEDS Intermodal Combination Responses None 2 Truck/Rail 29 Truck/Airport 21 Truck/Marine Port 24 Rail/Marine Port 21 Rail/Airport 11 Other 2 Table 21 describes the perceived shortcomings of the data. The most commonly cited shortcoming in several data sets was insufficient detail or inappropriate scale (25 respon- dents). Other common shortcomings included high cost (21 respondents), incomplete coverage of freight mode, move- ment, or commodity that is carried (19), datedness of the data (17), and small sample size and incomplete geographi- cal coverage (16 responses each). Eight respondents noted that the data had been developed for another purpose and were not adaptable, and five respondents indicated that the definitions were not applicable to their needs. Other short- comings (one citation each) were the “headquarter effect” associated with the PIERS data (i.e., the data reflected an administrative office location, not the location of the actual freight activity), the inadequacy of the reporting tool, the use of consultant resources (i.e., the knowledge base, if not also the actual database, is not in house), and not all of the data were compiled.

33 size, improve data quality, and reduce intrusion and respon- dent burden has attracted considerable attention in the freight planning community. Section 5 of the survey examined the use of ITS for freight surveys. Twenty respondents indicated that they used ITS to collect freight data. TABLE 20 PURPOSE FOR USING PUBLIC OR COMMERCIAL DATA SOURCES Purpose of Use Responses Land Use Planning 9 Infrastructure/Facility Planning 30 Traffic Safety Operations 10 Demand Management 14 Air Quality Management 6 Logistics Planning 9 Other 9 TABLE 21 SHORTCOMINGS OF AVAILABLE DATA Shortcomings of Available Data Responses Sample size/number of samples too small 16 Incomplete geographical coverage 16 Incomplete coverage of freight mode, movement, or commodity that is carried 19 Out of date 17 Insufficient detail or inappropriate scale 25 Developed for another purpose and cannot adapt to my needs 8 Cost is too high 21 Definitions not applicable to my needs 5 Other 3 Table 22 shows weigh-in-motion (WIM) technologies at 15 responses and sensors at 12 responses were the most com- monly used ITS technologies, followed by automatic vehicle classification (AVC) at 9 responses. AVC and sensors had the highest potential for integration with other data collec- tion initiatives (with 5 and 6 ratings respectively of “high” potential) followed by WIM (4 rated as “high” and 7 rated as “medium”). Six of the 20 respondents indicated that they currently link freight survey data with data from informatics such as roadway, on-board vehicle, and/or wide area sensors that can provide data on flows by mode, location, routing, and time of day. Eleven respondents did not make these linkages, two did not know, and one did not answer this question. TABLE 22 USAGE AND QUALITY OF ITS TECHNOLOGIES ITS Technology Do You Use? What Is the Potential for Integration with Other Data Collection Initiatives? Yes High Medium Low Advanced video image processing 2 1 0 1 Aerial videos 2 1 1 1 Automated vehicle classifi- cation (AVC) 9 5 2 2 Automatic vehicle identifi- cation (AVI) technologies 2 0 1 2 Automatic vehicle location (AVL) system 1 0 0 1 Cellular phone coordinates (probe vehicles) 2 1 1 1 Closed-circuit cameras (CCTV) 4 0 3 2 Electronic toll collection equipment 4 0 3 2 Environmental sensor stations 1 0 1 1 Global positioning system (GPS) equipment 3 1 1 0 License plate matching systems 4 2 1 2 Radio frequency identification 0 0 0 1 Smart cards 1 0 1 1 Vehicle tracking and navi- gation systems (VT&NS) 1 1 0 1 Weigh-in-motion (WIM) technologies 15 4 7 1 Sensors (e.g., loop detec- tors, acoustic sensors, infra- red sensors, and radar/ microwave sensors) 12 6 4 2 Respondents expressed both benefits to linking freight survey data with data from informatics and barriers to mak- ing these linkages. Table 23 lists the benefits. It should be noted that all respondents found some benefit to making these linkages. There were 10 citations each in having the benefit of improved data validation/quality control, increased accu- racy and data quality, and more comprehensive data. Nine respondents cited greater cost-effectiveness in data collec- tion, eight noted the reduced time between data collection and their availability, and six noted a reduced need for sur- veys and other data collection. Finally, there was one cita- tion each of the benefit accruing from having the informatics

34 data support planning efforts for surveys, getting a “bigger picture of what is moving,” and allowing for off system data collection. TABLE 23 BENEFITS OF LINKING FREIGHT SURVEY DATA WITH DATA FROM INFORMATICS Benefit Responses No benefits 0 More cost-effective data collection 9 Reduced need for surveys and other data col- lection efforts 6 Improved data validation/quality control 10 Increased accuracy/quality of data 10 More comprehensive data 10 Reduced delay between time of data collection and when available for analysis 8 Other 3 Table 24 lists the barriers to linking freight survey data with data from informatics. Fifteen respondents cited an insufficiency of capital resources to build the informatics infrastructure, and this was common to several data sets. An insufficiency in technical knowledge was cited 8 times, and a lack of standards for design and operations was cited 7 times. Other barriers (cited once each) were insufficient staffing, insufficient agency coordination, reluctance on the part of key partners to participate (in this case, a terminal operator), and a combination of lack of funds, internal infor- mation technology limitations, and institutional issues. TABLE 24 BARRIERS TO LINKING FREIGHT SURVEY DATA WITH DATA FROM INFORMATICS Barrier Responses Insufficient capital resources to build the informatics infrastructure 15 Insufficient technical knowledge to imple- ment or operate informatics 8 Lack of standards for design and operation 7 Insufficient staffing 4 Finally, one agency noted that it was planning to use more informatics. However, another noted some practical limitations with the use of ITS data: “there [are] limited ITS data OR the quality is too low for the purpose of plan- ning (i.e., Weigh-In-Motion data),” and a third respondent noted that “additional Weigh-In-Motion equipment would be helpful, but [the] technology has not met field condition requirements.” USER ASSESSMENT OF DATA Section 6 of the survey asked respondents to assess their data, whether sourced in-house or from external public or commercial databases. Respondents were also given the opportunity to comment extensively in their assessments. The results are summarized here. Assessment of Freight Data Table 25 summarizes respondents’ assessments of how well their freight data met their needs, which were expressed in six categories. The table also notes the applicability of the data to their activities. Key findings were as follows: Each of the six categories was applicable to respon-• dents’ needs, albeit to different degrees. Forecasting was the most applicable, at 37 citations (and four other respondents indicated that this category was not appli- cable). Investment decision making (31 citations), cost-benefit analysis (30), and operational analysis (26) were next, followed by design and environmental assessment (18 each). Most respondents indicated that the data met the needs • adequately (63 citations). A smaller number indicated that the needs were met poorly (46 citations) or “good” (36 citations). Eight described their needs as being met poorly (three for cost-benefit analysis) and seven described the data as being “very good” in meeting the needs. Needed Improvements Table 26 lists the improvements that respondents believed were needed to address current deficiencies and gaps. The respondents ranked the listed improvements on a scale of 1 to 5 (most important to least important). The two most important improvements identified were to provide more detail (17 responses) and ensure that data are collected regu- larly (14 responses). Other most important improvements included the collection of data that were not otherwise col- lected, ensuring the timeliness of the data (“[should not use] 2002 and 2005 data to make decisions in 2009, 2010 and 2011”), and management commitment and availability of capable staff. Respondents also identified several benefits to hav- ing these improvements. Thirty respondents indicated that these improvements would provide them with new capabili- ties, and 24 expected to benefit from improved productivity. (Respondents were allowed to provide multiple answers to this question.) Other specific improvements comprised bet- ter decisions and investments, faster identification of trends, better informed planning process, data that represent “what is moving,” [the identification of] economic opportunities, and more accurate freight counts and fees (one citation each).

35 “Willingness” of participants (in this case, termi-• nal operators) “to release accurate information” and obtaining much needed private data to fill gaps. Adequacy of the responses from the survey partici-• pants. Also important was the “interest in the [freight] community.” Level of specifics and detail in the responses.• Experience of the data collection team.• Timeliness and currency of the data: “freight informa-• tion needs to be current due to the rapidly changing economic conditions.” Expansion and Extension of Existing Data Collection Table 27 lists respondents’ interest in expanding or extending existing data collection, and also the priorities they assigned to these plans. With the exception of traffic counts (28 cita- tions), such plans were in the minority, although there was strong interest in expanding personal interviews (16 cita- tions, compared with 17 that were not being expanded or extended), focus and stakeholder groups (17 and 19 respec- tively) and GPS vehicle tracking (11 and 16 respectively). TABLE 25 ASSESSMENT OF HOW WELL NEEDS ARE MET BY FREIGHT DATA User Need Applicable to Your Activities? Degree to Which Your Needs Are Met? Yes No Very Good (exceeds your requirements) Good Adequate Poor Very Poor (not usable for your requirements) Forecasting 37 4 2 9 16 9 1 Cost-benefit Analysis 30 7 2 9 9 8 3 Operational Analysis 26 8 0 6 10 8 2 Design 18 11 2 1 8 5 1 Environmental Assessment 18 10 1 2 6 7 0 Investment Decision Making 31 5 0 9 14 9 1 Note: Responses from two different offices at the California DOT were included in this table. TABLE 26 IMPROVEMENTS NEEDED TO ADDRESS DEFICIENCIES OR GAPS Improvement Frequency of Rank of Importance 1 (most important) 2 3 4 5 (least important) Expand Sample Size 5 9 3 8 8 Expand Coverage 5 3 13 5 6 Expand Modes 8 3 8 7 6 Provide More Detail 17 9 2 4 4 Ensure Data Are Collected Regularly 14 9 3 5 2 Other 4 0 0 0 1 Note: Responses from two different offices at the California DOT were included in this table. Success Factors in Data Collection Respondents identified several factors for success in their collection of freight data: Adequacy of funding was the most dominant theme. • Funding was needed “to pay for a sophisticated freight forecast product. Most MPOs do not have the resources.” One respondent noted that freight data are “highly desired,” so it has been able “to pay for what is out there.” Prior knowledge and experience in the subject were • cited as success factors, including “knowledge of sur- vey techniques [and] knowledge of freight market” and “many years of experience in this type of data collec- tion, analysis and reporting.” Appropriate planning of the data collection effort, • clarity in objectives (“focus on what problem needs to be solved”), clarity “in the questions that can be answered,” and “asking the right questions.” Effective communications and “relationships” with the • survey participants.

36 Cost relative to benefits: “why pay for large surveys • when there are very limited funds to invest?” Data quality. One respondent cited the potential for • underreporting as a problem: “The data [are] self-re- ported by terminal operators through an honor system. There is the potential for underreporting as harbor fees are associated [with] freight movement.” Irregularity of data collection and data gaps.• One respondent compared different levels of data and listed the associated difficulties: “Facility operation data [are] collected internally and [are] generally of high qual- ity, but [are collected] too infrequently due to budget con- straints. In contrast, regional and national data [are] not satisfactory for many purposes, especially intra-regional/ local freight flows and characteristics. Data access restric- tions due to bureaucracy. Little consistency in data format- ting makes [them] difficult to process. Would welcome open platform for data sharing.” Another respondent encapsulated the complexity of the needs by noting that the data are “not current, not geographically detailed or [temporally] detailed, no historical trend [information] on change and paradigm shifts, expensive to collect, difficult to repurpose, inevitably not what you need but what you have.” Technical and Content Problems and Limitations Respondents addressed a series of technical and content problems and limitations that they experienced with pre- vious surveys, and described how they planned to address TABLE 27 INTEREST IN EXPANDING OR EXTENDING EXISTING DATA COLLECTION Type of Survey/Data Is the Existing Data Collection Effort Being Expanded? If Yes, Please Indicate Level of Priority Yes No or N/A Low Moderate High Roadside/Intercept Survey 9 27 2 4 5 Combined Telephone Mail-out/ Mail-back Survey 6 25 0 3 2 Telephone Survey 5 25 2 2 1 Mail-out/Mail-back Survey 5 24 2 3 0 Personal Interviews 16 17 4 7 2 Internet Surveys 4 23 2 3 1 Focus and Stakeholder Groups 17 19 3 6 6 Administrative Data (e.g., insurance records) 3 23 1 0 1 Commercial Vehicle Trip Diaries (e.g., trip logs) 5 22 2 1 2 GPS Vehicle Tracking 11 16 4 3 3 License Plate Match—Manual 4 23 3 2 0 License Plate Match—Electronic 3 23 2 0 0 Traffic Counts 28 10 3 13 7 Problems with Existing Data and Most Important Improvements Respondents were asked to list the main problems with their existing data and to indicate what they saw as the most important improvements needed. The following were key points: Greater detail and disaggregation of existing data: • “Commodity data from [Bureau of Transportation Statistics] or [TRANSEARCH are] at too high a level.” There is “no level of detail for an area the size of a [metropolitan region].” “Need zip code or [traffic anal- ysis zone] level of data.” “Would like more detailed O-D [information].” “Need for … more specific data.” “[Need] greater geographic detail.” Improved coverage and “much bigger sample size”• Improved response rates• More timely, up-to-date data. “Public data, like FAF2, • [are] too old.” Origin-destination data• Travel delay data• Routing data—“not just shortest path.”• “Capacity to capture the data. [State DOT does not] • have the staff or capacity and [so the State DOT uses] consultants which are too costly.” “Need some dedi- cated staff working on data.” “The main problems with existing data [are] knowing what information is pro- vided in the source and also having the staff with the expertise to utilize the data.”

37 them in future surveys. Seven categories were specified in the survey; the findings are summarized here: Problems of precision. Responses1. : hiring the “best consultants” to assist “on the forecast,” requiring bet- ter calibration of data collection equipment, increased funding, better training of temporary staff “to ensure accuracy of data,” standardized survey instruments, use of current technology, and working with another public agency to secure the required data (in this case, ship manifests). Problems of confidentiality. 2. Responses: increase funding to “buy better [information],” focus only on the data that are “essential” and “stop wasting … time and money on data that the industry doesn’t support in [these] settings,” work with data “owners … to change restrictive policies when confidentiality can be assured,” assure confidentiality to respondents, and, explain the cost [i.e., the implications] to the gov- erning legislature. Problems of misunderstood survey questions. 3. Responses: explaining the “benefit” of the surveys in ways that are meaningful to respondents, develop “better” plans for surveying (“in-house surveying? Contract … QUALIFIED data collectors.”), improve questions on mail-out and internet surveys, use a consultant to develop the survey, conduct pilot tests, provide assurance that the agency “will aggregate the data [so as not to] reveal information about their company but they are still concerned about sharing the information,” improved staff training (“train the trainer”), provide capabilities in multiple languages, provided “detailed and helpful route graphics,” and modify questions. Problems of unintended applications4. . Response: need to address “translation and cross cultural challenges.” Problems of malfunctioning equipment5. . Responses: experimenting with image data collection, rapid repairs, and responses. Problems of low response rates6. . Responses: consider more concise surveys, compose better cover let- ters, provide incentives, allow “multiple avenues for response, e.g., internet, 800 telephone number,” use multilingual survey instruments, reduce survey dura- tion, follow up with telephone calls (to address low response “in rural areas”), increase funding and hire a better consultant. Other problems. (This category allowed respondents 7. to identify specific issues not otherwise addressed.) Responses: “some potential respondents who agreed to be interviewed for the survey initially, refused to take the survey later on. A proper [redress] of their questions/concerns can improve sample size,” need to address “concerns about how the data will be used and revealed to the public.” Legal and Confidentiality Issues Finally, respondents were asked to describe how legal and confidentiality issues have impacted the design and conduct of their surveys, and steps they have taken to address these issues. The challenges comprised— The level of detail for [analyzing or presenting] origin-• destination and commodity information is impacted by confidentiality concerns. Some participants (motor carriers) are unwilling to share • information, such as that regarding commodities. Attempts to expand exemption of the completed sur-• veys from the Freedom of Information Act have been unsuccessful. Use of [some] proprietary data can be limited by confi-• dentiality constraints. Surveyors at international border crossings were pro-• hibited by border and customs staff from conducting interviews in the queues (meaning that another survey method had to be found). Solutions to these challenges comprised— Inclusion of a legal review with all (of one respon-• dent’s) surveys. Development of appropriate wording regarding legal • and confidentiality conditions: “Increases [the] time to develop contracts, but have found that challenges are not prohibitive.” Avoidance of the “level of detail” [in surveys] that • would generate legal and confidentiality problems. Conduct of all correspondence with survey participants • via a third-party customer service center—in this case, with electronic toll tag owners. Continued strengthening of the relationship with cus-• tomers “to create trust.” Not collecting addresses.• Allowing for voluntary participation in the survey.•