Govindan Kannan

Planning and estimating earthmoving operations rely on the ability to collect field data and analyze operations. Two recent developments have dramatically changed the methods used and the results obtained in this process. The first relates to data collection where instrumentation placed on loading and hauling units can replace stopwatches as a primary source of data. The second relates to analysis where simulation techniques can replace deterministic job studies. This paper reviews these developments and shows that data collection and analysis techniques must complement each other. The use of stopwatches may satisfy the requirements of job studies but does not satisfy those of simulation. Simulation requires knowledge of the variance associated with the activity duration. Obtaining it from a few data points collected using stopwatches is frequently statistically incomplete. Simulation requires a continuous form of data collection to support its modeling process and, hence, offers new challenges to the data collection methods. Instrumented vehicles present a viable opportunity to collect continuous field data. However, the data collected using the instrumented vehicles must be characterized in such a way to reflect the actual job conditions under which they were recorded. As a result, this research extends the concept of data collection to building an experience database. Field data and visualization techniques are presented in this paper to illustrate the concept of the experience database.

Planning and estimating earthmoving operations is a challenging assignment because of the complexities and the uncertainties involved. The analysis during planning and estimating stages can be accomplished through deterministic or stochastic methods with data provided by manual or automated data collection technologies. It can be argued that the proficiency of each analysis methodology is often dependent on the characteristics of the corresponding data collection technology. Electronic sensors, load cells, and microprocessors — collectively referred to as vehicle instrumentation — have made it possible to automatically collect field cycle time data for earthmoving operations. Although automatic data collection overcomes several drawbacks associated with manual data collection, it posses challenges of a different nature. The primary challenges are the need to handle a large volume of indiscriminate data and the need to associate the field data with the operating conditions. The methodology to handle the large volume data is discussed in Kannan et al. This paper addresses the issue of associating the cycle time data and the corresponding operating conditions through the use of images. The process of identifying the operating conditions using images and linking it to cycle time data collected by vehicle instrumentation is termed as characterization.