Recent Doctoral Dissertations Ambo, Alemayehu. The University of New Brunswick (Canada), 1990. This research focused on the consideration of sample highway sections; the estimation of traffic measures; the estimation of relevant highway costs; and finally, the allocation of the costs to the vehicles recorded during the traffic count survey. The major source of the traffic data used in the research was the University of New Brunswick Transportation Group 1987 traffic count survey conducted along the Trans Canada Highway on sections at St. Leonard, Woodstock, Coles Island, Sussex, Shediac, and Aulac. The vehicle classifications used in the research were: passenger cars; light trucks and vans; recreational vehicles; buses; 2-axle, 3-axle, and 4-axle single unit trucks; 5-axle, 6-axle, and 7-axle transport trucks; and pups. Using these traffic data and information on the 1984 truck origin-destination surveys taken in the Atlantic provinces; traffic measures, i.e. volume of vehicles, passenger car equivalents, and equivalent standard axle loads were estimated. The accounting period was assumed to be 60 years, indicating three design periods and three traffic growth scenarios. The costs considered for the replacement of the existing pavement and the future maintenance of the highway sections were current costs provided by the New Brunswick Department of Transportation. The major costs considered were: roadway, shoulder, maintenance, right-of-way, future overlays, traffic control devices, etc. The costs were attributed to the traffic measures and then costs per vehicle per kilometer were estimated using different costing methodologies. The research revealed insignificant cost differences between passenger cars, and light trucks and vans. It also showed that using different design periods did not produce significant cost differences except when different traffic growth rates were applied. The research also produced findings that indicated that conventional costing methodologies could enable the highway agency to collect more than its expenditures; therefore, the proper approach could be to charge the vehicles on a year-by-year basis which closely matches the public accounting procedures used by municipal, provincial and federal authorities which construct and maintain roads and streets on a continuing basis. The research also produced results which justify the consideration of using different highway sections and truck weight weigh station data during the cost allocation exercise.
Truck weight prediction modeling. Nassiri, Habibollah Shirabad. Texas A&M University, 1989. 291 pp. Chair: Donald L. Woods. Order Number DA9015554 A reasonable forecast of anticipated traffic loadings
is necessary to design or rehabilitate a highway. Underestimating the load experience for a highway could result in an under-designed facility, leading to the need for major unanticipated repairs. An overestimate, however, could result in the construction of over-designed facilities, tying up money badly needed for other projects. The results of this study should be helpful in providing a better understanding of prediction of truck loads and their trends for future decision making processes. This dissertation considers the single axle, tandem axle, and gross weight distributions of seven truck types at three weigh-in-motion (WIM) stations in Texas between 1977 and 1985. It also provides evaluation of the trend in average weights over this nine year period. In addition, predictability of future truck weight distributions based on previously collected data are examined. Results cast doubt on the concept of statistically sampling truck weights at a site and projecting future trends in the axle load distribution at a particular site. The results suggest that predicting axle loads at another site base and sampling data on roadways of similar classifications may not be possible with any degree of statistical reliability.
PROJECT M A N A G E M E N T
Management of the pre-construction process for highway projects using manpower demand prediction models. Persad, Khali Ram. The University of Texas at Austin, 1989. 375 pp. Supervisor: James T. O'Connor.
Order Number DA9016954 The design of a comprehensive microcomputerbased system for managing pre-construction activities and programs in a State Department of Highways is described. Project-related activities from conception to award of a construction contract are researched and structured. Performance on a large sample of projects is analyzed, and models for predicting engineering manhours and durations are developed. Alternative models for estimating manpower demand across a long-term program are presented. The results are integrated in a modularized management system for planning and control of pre-construction projects and programs. This is the first identified formal investigation of the requirements for managing the pre-construction process. Managerial experience is captured and systematized. The trade-offs between organizational needs and project objectives are identified in the system design. Models for forecasting engineering performance are quantified. The results advance current knowledge of management of engineering activities.
Recent Doctoral Dissertations
Measuring the quantitative and qualitative im- highway construction projects in the State of pact of the Disadvantaged Business Enterprise Florida. requirement of the Federal Surface Transportation Assistance Act as implemented by the Wisconsin Department of Transportation. Manning, David. The University of Wisconsin-Milwau- Strategic development of transport systems: A study of the physical constraints on planning kee, 1990. 231 pp. Supervisor: Frank Besag. Order Number DA9035147 processes. Anderson, David Laurence. Aston University (United Kingdom), 1987. 394 pp. Order Number BRDXg961g The purpose of this study was to determine the qualitative and quantitative impact of the implementation of the Disadvantaged Business Enterprise (DBE) Program by the Wisconsin Department of Transportation. The program was a requirement of the Surface Transportation Assistance Act passed by Congress in 1982. A quantitative analysis was conducted by comparing the two 4-year periods of 1980-83 and 1984-87. The study also used a questionnaire survey to do a qualitative analysis to measure the perceived impact of the program by the participants involved. The study showed that the DBE program implementation resulted in: (i) significant increases in the dollar amount spent with disadvantaged firms, (ii) a major increase in the number of contracts DBE firms received from the Department of Transportation, and (iii) a significant increase in the percent of transportation dollars being spent with disadvantaged businesses. The study also found that program participants perceived that the program was effective in meeting many of their business needs by increasing opportunities for their firms to do business with the Department.
Model for prediction of highway construction production rates. Ellis, Ralph Donald, Jr. University of Florida, 1989. 272 pp. Chairman: Zohar Herbsman.
Order Number DA9021846 Construction production rates are important in many construction management functions. Cost estimating, cost control, scheduling and resource allocation all rely upon production rate data. The prediction of future production rates is essential. Construction production rates are extremely variable. Historically, the prediction of production rates has been difficult and often inaccurate. This dissertation presents the concept of a factorial model for explaining the variance associated with construction prediction rates. Production rates are affected by many influencing factors. Identification and quantification of the influencing factors allow a more complete understanding of the work process. The considerations of factor identification and data collection systems are discussed. Model development and statistical procedures are presented. A comprehensive approach for developing a production rate prediction model is developed. A demonstration prediction model is developed from a survey data base of production rate observations taken from 60 different
Investment in transport infrastructure can be highly sensitive to uncertainty. The scale and lead time of strategic transport programs are such that they require continuing policy support and accurate forecasting. Delay, cost escalation and abandonment of projects often result if these conditions are not present. The conventional treatment of uncertainty is a development of the "demand satisfaction" approach; the emphasis is on the forecasting of demand and those solutions which are most robust to the possible futures defined are preferred in scheme appraisal. In this study, the emphasis is on the possible solutions. It is assumed that the future is inherently uncertain and the requirement is for projects which reduce the sensitivity to this uncertainty. In Part One, the characteristics of infrastructure such as scale and lead time are identified as significant contributors to this sensitivity and as major constraints on planning processes. The extent to which current strategies and techniques acknowledge these constraints is examined. In Part Two, a simple simulation model is developed to evaluate the effects of these constraints. The model is used to assess the importance of scale and lead time in two major projects: the third London airport and the development of the London road network. In conclusion, the scale of infrastructure investment rather than its lead time is considered the most important of the constraints on the processes of transport planning under uncertainty. Adequate appraisal of such constraints may best be achieved by evaluation more closely aligned to policy objectives.
Analytic optimization of bus systems in heterogeneous environments. Chang, Shyue Koong. University of Maryland College Park, 1990. 172 pp. Director: Paul M. Schonfeld.
Order Number DA9030873 Analytic models for optimizing transportation systems have typically sacrificed geographic details and temporal variations in demand and supply parameters in order to identify optimal relations among certain decision variables and parameters in closed form. The applicability of such models has been limited by their simplifying assumptions about idealized