Degree Name

Master of Total Quality Management (Hons.)


This thesis addresses the task of reducing the risk to the Roads and Traffic Authority of NSW (RTA) of accepting poor quality work from its suppliers by the implementation of an effective statistical compliance scheme. Further, the compliance scheme is designed to lead to the introduction of statistical process control into the road construction industry. This will be achieved by control charting the statistical summaries required for the compliance scheme and by requiring the contractor to demonstrate statistical process control. Therefore, this thesis also highlights the importance of measurement, and analyses of measurement, for quality improvement.

The quality characteristic tested for acceptance by the compliance scheme is the difference between the constructed height of the pavement course surface and the design height of the pavement course surface. Control of the surface height of the pavement course controls the pavement course thickness.

Statisticians have noted that the assumption that processes generate independent and identically distributed random variables is not always appropriate for quality control purposes. This thesis estimates the effect of data correlation on surface level departure and thickness measurements by time-series analysis. This enabled a process capability analysis of the processes for constructing concrete pavement to the correct height and thickness, and the preparation of a compliance scheme for pavement surfaces.

The application of statistical techniques allows for an equitable sharing of the risks, the compliance scheme, between the RTA and road construction contractors. The effect of this is, and the change of focus to process measurements, are expected to support a cultural change to more of a "partnering relationship" between the RTA and contractors.

Compliance schemes are limited by the capability of the measurement system used. For this reason, analyses were also carried out on the measurement system (survey) that currently measures surface level departures. It is concluded from these analyses that survey measurements by EDM trigonometrical heighting are capable of measuring the process to an accuracy that can track process improvement. It is also concluded that concrete pavement course thickness measurements by survey are substantially better than the current method of edge measurements and core depths.

The ability of the general surveying profession to measure the process to the required accuracy was estimated by a questionnaire distributed to surveyors throughout N S W involved in road pavement construction. The questionnaire was also to determine the adequacy of process controls for constructing pavement surfaces to the correct height. Responses to the questionnaire determined that few surveying procedures were adequately controlling all the factors that affect the accuracy of surface level departure measurements. Some process controls and surveying techniques were also identified as having an adverse effect on quality.

The road pavement is the most dominant influence on the long term cost of road asset management because of its high construction and maintenance costs. The R T A Pavements Manual (1993) estimates that concrete base courses constructed 1 0 % less than their design thickness will reduce the design life of road pavements by approximately 9 0 % . In 1996/97 the R T A spent approximately $297million on maintenance and rehabilitation of pavements under its control and an additional amount of about $200million on pavement construction. The change of focus to process measurement is expected to have a significant positive impact on the RTA's most costly asset.

The compliance scheme will, for the first time, provide an accurate measure of constructed base course thickness to assist asset management. This will also aid research into pavement design and assist problem solving in the event of premature pavement failure. In addition, by measuring contractors' processes, the compliance scheme will provide a performance indicator to assist contract administration.



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.