The present paper demonstrates the applicability of genetic algorithms, as an optimization tool capable of overcoming combinatorial explosion, to the road‐maintenance planning problem at the network level. Genetic algorithms are search algorithms based upon the principles of Darwinian evolution. The concept of the survival of the fittest is used in a structured, yet randomized, information exchange to form a robust search algorithm. Genetic algorithms efficiently exploit historical information to locate search points with improved performance. The theoretical basis and operations of genetic algorithms are presented. A computer model, PAVENET, formulated on the operating principles of genetic algorithms to serve as an analytical aid for pavement maintenance engineers, is introduced. The formulation of the PAVENET model is described in detail. Analyses are conducted to show the characteristics of important operating parameters of the PAVENET program. These parameters include: (1) Parent pool size; (2) mutation rate in offspring generation; and (3) ranking system for offspring selection. The convergence process of a sample problem as analyzed by the PAVENET program is studied and recommendations on the choice of operating parameters are made.

This paper analyzes the road‐maintenance planning problem at network level using a genetic‐algorithm program, PAVENET, which is a computer model developed to serve as an analytical tool for pavement maintenance engineers. The effects on road‐maintenance programs of the following parameters are analyzed: (1) Network parameters, such as the age structure of pavements; (2) maintenance‐policy parameters, such as the choice of warning levels; and (3) resource parameters, such as the level of budget available. Examples are presented to illustrate how the pavement age structure of a road network affects road‐maintenance planning, and how the PAVENET program can be used to formulate a maintenance strategy to regulate long‐term maintenance demand. Also illustrated are the adverse long‐term effects of poor budget planning in the initial years. Analyses are presented to demonstrate the application of the PAVENET program to the selection of maintenance warning levels. Finally, examples are given to study the effect of the type of pavement and distress function on maintenance planning.

The major objective of this research was to apply a probabilistic procedure to evaluate the safety effectiveness of pavement marking of undivided rural roads. A 100 sites were randomly selected from the population of improved roads in the state of Indiana. In the analysis, accident reduction factors at the individual site level were estimated as a proportion based on the expected accident rates. To eliminate the effect of regression to the mean, a Bayesian approach was adopted to estimate expected accident rates in the before and after periods. Results of the analysis indicated that the beta distribution adequately fitted accident reduction factor frequencies. Also, accident reduction factors at individual site level were found to be positively correlated with accident rates in the before period. When pavement marking was considered at all sites irrespective of their accident experience, no statistically significant safety impact was observed. However, pavement marking at hazardous sites provided a significant level of accident reduction. Thus, it can be concluded that pavement marking does improve the safety of hazardous sites on rural undivided highways.

The effect of restricting vision on driving performance was investigated in a field study. Commonly occurring binocular visual‐field defects were simulated for a group of young normal subjects and the effect of these defects on their driving performance on a private closed rural road, free of other vehicles, was assessed. The monocular condition did not significantly affect performance for any of the driving tasks assessed. Restriction of the binocular visual field to 40° and less, significantly increased the time taken to complete the course, reduced the ability to detect and correctly identify road signs, avoid obstacles, and to maneuver through limited spaces. Accuracy of road positioning and reversing were also impaired with field restriction. However, the time taken for many driving tasks, reversing and maneuvering, and the driver's ability to estimate speed, stopping distance, or reaction time were not affected by a restriction of the binocular visual fields. The results are discussed with regard to their impact on traffic engineering practices.

To date, efforts to manage freeway congestion have been seriously impeded by the inability to promptly and reliably detect the presence of traffic incidents. Traditional incident‐detection algorithms distinguish between congested and uncongested operation by comparing measured traffic‐stream parameters with predefined threshold values. Given the range of possible operating conditions in the traffic stream, selecting a single threshold value, and the suitability of that selected threshold, is full of uncertainty. This inherent uncertainty makes fuzzy logic a promising approach to incident detection. A Fuzzy Logic Incident Patrol System (FLIPS) is proposed to solve many of the problems inherent in traditional incident‐detection algorithms. The FLIPS combines fuzzy logic with the learning capabilities of neural networks to form a connectionist model. The system can be constructed automatically from training examples to find the optimal input/output membership functions. Threshold values, implicitly obtained by fuzzy‐logic rules and membership functions, are treated as dependent variables, which change according to prevailing traffic‐stream parameters measured by detectors. The FLIPS avoids the rule‐matching time of the inference engine in the traditional fuzzy‐logic system. The potential effectiveness of the FLIPS is evaluated using an empirical database collected in Toronto, Canada. Future refinement to the FLIPS are also discussed in this paper.

A model based on an expected total‐cost minimization approach is developed for determining the optimal‐degree curvature of simple horizontal curves on undivided two‐lane highways. Expected total cost is defined as the sum of the construction costs and the expected cost of accidents. Construction cost is treated as a deterministic function of curvature, and the expected cost of accidents is determined in relation to the amount by which required (demand) curvature differs from design (supply) curvature. It is shown that the optimal degree of curvature is highly sensitive to the skewness of the probability distribution of the required curvature. This distribution is derived from the fundamental relationship between speed and degree of curvature, treating speed to be a random variable. The optimal value is shown to change in direct proportion to skewness of the speed distribution. User cost, in terms of travel time, is treated as a special case when examining the sensitivity of optimal design curvature to construction‐ and accident‐cost functions. Model applications in curve design are illustrated by a numerical example.

A procedure is described that can be used to validate a knowledge‐based expert system. The method applies four qualitative measures that are specially designed for problems solved by multiple levels of experts whose authority and expertise can be modeled hierarchically. Development of test cases and performance measures are discussed. The proposed method is applied to a system that selects and ranks bridge rails from a list of candidate configurations. This rule‐based system aids engineers in selecting bridge rails according to conditions of the roadway, bridge, and surrounding environment. Based on the level of agreement with human experts and the Turing test, it is shown that the prototype system exhibits a relatively high level of expertise. Other tests and observations indicate that some system deficiencies exist and several features need modification and refinement.

Rail continues to be an important transportation mode. Like other transportation infrastructure, maintenance and replacement are critical to ensure safe, efficient operation. Research and development of technologies for condition‐data collection and computer‐based decision‐support systems are not widely reported in the literature due to the nature of the industry. This paper presents an overview of technologies for railroad track‐condition assessment and decision‐support systems for track maintenance and renewal. Track inspection methods range from visual inspection to the use of sensors and sophisticated measuring systems for the identification of corrugation, measurement of track geometry, measurement of rail wear, and location of internal rail flaws. The data produced by these systems may be integrated into a track‐quality index, or a component index, and aggregated by segment. These data are then used for safety evaluations, work planning, and strategic planning of repair and replacement. Decision‐support systems attempt to move maintenance and replacement decision making from an art to a science. They are widely used by the larger railroads. A variety of systems and issues are briefly described.

This paper presents the formulation and the application of a continuum‐based finite‐layer approach to evaluate dynamic pavement strains under moving traffic load. The dynamic material properties (complex shear modulus and Poisson's ratio) of the asphalt‐concrete (AC) layer can be varied as a function of the loading frequency. The predictive capability of the model is illustrated through a parametric study in which two typical thin and thick pavement sections were subjected to a semitrailer moving at different speeds. The results are very similar to those reported earlier from a field‐testing program. The results show that the pavement strains are strong functions of the pavement thickness and the vehicle speed, and they reduce substantially with the increase in the speed of the vehicle. Also, a substantial compressive‐strain component is present at the bottom of AC layer that may be important for any realistic fatigue‐life estimation of the pavement. The ratio between the maximum compressive strain to the maximum tensile strain at the bottom of the AC layer can be as high as 1.01 for the thick pavements under tandem axles.

Congestion of aircraft traffic along airways and in the terminal areas of large airports has significantly increased in recent years. To handle the new levels of congestion in the skies a better understanding of air conflicts is necessary. This paper presents a stochastic model that evaluates the number of conflicts for a spatial geometry of two intersecting airways. The stochastic nature of any desired parameter that defines aircraft location or performance characteristics is accounted for by the simulation process. The structure of the model ensures its efficiency in terms of computer resources as well as its accuracy. The effect of shape of the arrival distribution functions on the shape of the conflict distribution function and its variance was investigated. The variance of the number of conflicts is dependent on the aircraft‐arrival distribution. This association is not a simple one, since an airplane on one path can conflict with more than one airplane on the other path. Thus the covariances in the variance term are dependent on the interarrival time described by the distribution. In a similar way the shape of the distribution function of conflicts was found to be dependent on the aircraft‐arrival rate distributions.