【正文】 trut pressure, gear control lever position,bed raise switch position, and speed to derive the duration of truck cycle ponents. Truck cycle ponents or activities are defined in terms of the parameters. It is evident from the literature reviewed that traditional datacollection methods are limited by the ability of an observer to instantaneously decide activity start and stop times within a narrow and static field of view. Modern techniques remove this limitation by using sensorbased datacollection techniques to identify key times through changes in sensor output. A thorough understanding of the methods for reviewing, reducing, and analyzing GPS data from construction operations requires background knowledge of how and what data are collected.Data Capture An automated data capture system based on GPS technology was used to record the raw data at a userspecified time interval, including date, time, velocity, and horizontal position. The system consists of a data box, junction box, and sensor pack and was developed based on field evaluations of mercially available systems. The data acquisition and storage box is a weathertight enclosure that houses the system circuitry and the CompactFlash media on which the data is stored. The sensor pack consists of a Garmin GPS35 LVC receiver and a Vector 2X magnetic pass manufactured by Precision Navigation, Inc. In addition to the GPS and pass, the system has been developed such that additional analog and digital sensors can be incorporated into the system for future research. The junction box serves as a central location for all electrical connections to be made. Power is made available in the box at 5 V for any additional analog sensors incorporated into the system. Data are recorded in ma delimited ASCII text format, with a separate data file for each day of collection to ensure files of a manageable size. The parameters recorded by the system include unit identification number, date, UTM time, latitude, longitude,velocity, type of GPS fix, output from the five analog and eight digital sensors, and direction of travel from the pass. Preliminary ReviewA tradeoff exists between the volume of data recorded and the level of detail provided by the data. Large volumes of data are produced when recorded at a time interval sufficiently short to facilitate operations analysis. The several thousand records produced daily on each machine must be reduced to the relatively few key records necessary to calculate activity times. The large volume of data dictates that the data reduction process be automated.The format of the data, a text string repeatedly recorded at aset time interval, is conducive to analysis in a spreadsheet environment supported by capable graphics. Microsoft Excel was chosen as the data processing engine as it is a widely used and recognized program already familiar to many potential users, thereby minimizing or eliminating the learning curve. Excel is a mon and robust spreadsheet application that can be automated through the Visual Basic for Applications (VBA) language and provides the means to graphically represent data.Processing the collected data for evaluation and reduction is performed in three steps: the data into Excel。 and graphical representations of the data. These three steps have been automated using VBA code, and this routine is available from the system user interface, provided as Fig. 1. The process of importing, converting, and graphing is executed by selecting the Import Text option, which allows the user to point to the desired data file.Once the file is identified, the data are imported into Excel with each record in a separate row and each parameter in a separate column. The geodetic positions are recorded as latitude and longitude in units of degrees and minutes. The scale of even a very large construction earthmoving operation is sufficiently small to make geodetic coordinates less than meaningful, and therefore the recorded geodetic coordinates are converted to a planar coordinate system in units of meters. The methods to convert from geodetic to state plane coordinates are outlined in standard surveying texts (Stem 1990).In addition to importing the data and converting the coordinates, two separate graphs of the data are prepared to assist the user in reviewing and understanding the data. Mapping the recorded coordinates creates a plan view of vehicle position throughout the workday, and color is used to distinguish between positions recorded when velocity is below a userdefined value,most frequently zero, and those recorded when velocity was greater than the value. A plan view of data collected during one day of the field application is shown in Fig. 2。 spent with velocity equal to zero within a fixed distance of a moving location—., a haul truck loading at a loader。 position of the truck must be within a userdefined area。 and previous record was outside the area for enter times, or the next record is outside the area for exit times.The FBTIM first evaluates position to determine whether therecorded position is within defined boundaries. Then it determines whether the record represents an enter or exit time by examining the status of the previous or next record. The user defines the boundaries on the FBTIM input dialog box and may define as many as 4 areas by providing a minimum and maximum value for both northing and easting coordinates for each area. The values are used to evaluate each recorded position and mark it as in or out of each area. The times of entry and exit are then identified and summarized.Case StudyThe system was field tested by collecting data from three Caterpillar 771 quarry trucks and a 988 rubbertired loader used to move material at a local aggregate quarry. Shot rock was hauled to crushers, while overburden material was hauled to an onsite waste area. Three load areas and three dump areas were identified: the load areas were active rock faces