Knowledge-Based Scheduling Analysis Software for Highway Rehabilitation and Reconstruction Projects (05-0567)
Eul-Bum Lee, University of California, Berkeley
John Harvey, University of California, Davis
Michael M Samadian, California Department of Transportation

Most of California¡¯s highway infrastructure was built between 1955 and 1970. These pavements had 20-year design lives and many now require frequent maintenance. In 1998, the California Department of Transportation (Caltrans) launched the Long-Life Pavement Rehabilitation Strategies (LLPRS) program to rebuild approximately 2,800 lane-kilometer (lane-km) of high traffic volume urban freeway in the 78,000 lane-km state highway network over a 10-year period. Priorities identified for the successful implementation of LLPRS projects are the selection of construction schedules and the development of traffic management plans that minimize road user and agency costs. This paper presents a construction simulation program, called CA4PRS (Construction Analysis for Pavement Rehabilitation Strategies), developed as a scheduling and production analysis tool for LLPRS projects for use during the planning and design stages. CA4PRS estimates the optimized distance and duration of highway rehabilitation projects, taking into account the constraints of scheduling interfaces, pavement design, lane closure tactics, and contractor logistics. As a knowledge-based computer system on a Microsoft Access database, it uses Monte Carlo simulation, critical path method (CPM) analysis, and linear scheduling technique. CA4PRS is designed to help highway agencies and paving contractors make construction schedule decisions that balance rehabilitation productivity, traffic inconvenience, and agency cost. Application of the CA4PRS model to urban freeway rehabilitation projects in California, including the Interstate-10 (I-10) Pomona, Interstate-710 (I-710) Long Beach, and Interstate-15 (I-15) Devore projects, has demonstrated its value in saving millions of dollars for both Caltrans and road users.