Marina Marinelli

Earthmoving trucks condition level prediction using neural networks

Purpose – The purpose of this paper is to present an artificial neural network (ANN) model that predicts earthmoving trucks condition level using simple predictors; the models performance is compared to the respective predictive accuracy of the statistical method of discriminant analysis (DA). Design/methodology/approach – An ANN-based predictive model is developed. The condition level predictors selected are the capacity, age, kilometers travelled and maintenance level. The relevant data set was provided by two Greek construction companies and includes the characteristics of 126 earthmoving trucks. Findings – Data processing identifies a particularly strong connection of kilometers travelled and maintenance level with the earthmoving trucks condition level. Moreover, the validation process reveals that the predictive efficiency of the proposed ANN model is very high. Similar findings emerge from the application of DA to the same data set using the same predictors. Originality/value – Earthmoving trucks’...

Investigation of earthmoving trucks deterioration using Discriminant Analysis

Earthmoving equipment has a determinant role in the successful realisation of most civil engineering projects. However, it may suffer significant downtime due to the continuous and intense use in harsh working conditions. This downtime may be associated with certain deterioration parameters, which if known, would allow more accurate estimations to be made. For this purpose, the data sets from two large Greek construction companies containing the characteristics of 126 earthmoving trucks (capacity, age, kilometres travelled to date, maintenance class and condition level) have been analysed using discriminant analysis. The analysis allows for the assessment of the connection of each characteristic with the condition level of the sample trucks and also leads to rules that can be used for the prediction of the condition level of other trucks.

Algorithmic method for scraper load-time optimization

AbstractScrapers have established an important position in the earthmoving field as they are independently capable of accomplishing an earthmoving operation. Given that loading a scraper to its capacity does not entail its maximum production, optimizing the scraper’s loading time is an essential prerequisite for successful operations management. The relevant literature addresses the loading time optimization through a graphical method that is founded on the invalid assumption that the hauling time is independent of the load time. To correct this, a new algorithmic optimization method that incorporates the golden section search and the bisection algorithm is proposed. Comparison of the results derived from the proposed and the existing method demonstrates that the latter entails the systematic needless prolongation of the loading stage thus resulting in reduced hourly production and increased cost. Therefore, the proposed method achieves an improved modeling of scraper earthmoving operations and contribute...

Early Bill-of-Quantities Estimation of Concrete Road Bridges: An Artificial Intelligence-Based Application

Accurate cost estimation in the preliminary stages of project development is critical for making informed planning decisions. However, such early estimates are typically restricted by limited information. In this article, the widely recognized intelligence of feed-forward artificial neural networks (FFANNs) is used to process actual data from 68 concrete road bridges and provide a surrogate model for the accurate estimation of the bill-of-quantities (BoQ). Specifically, two FFANNs are trained to estimate the superstructure and piers concrete and steel-based on the construction method and the bridge dimensions. As the relevant metrics demonstrate, the FFANNs capture very well the complex interrelations in the data set and produce highly accurate estimates. Furthermore, their generalization capability is superior to the capability of respective linear regression models. As the data used to train the FFANNs are normally available early in the project lifecycle, the proposed model enables early, yet accurate cost estimates to be obtained.