Deep Neural Network Models for Improving Truck Productivity Prediction in Open-pit Mines

The accurate prediction of truck productivity plays a pivotal role in improving the efficiency and profitability of open-pit mining operations. However, predicting truck productivity is challenging owing to the complex nature of the working conditions of the mine site. This paper proposes a deep neural network model to overcome the challenge of predicting truck productivity in open-pit mines. The prediction model was built using eight variables and was optimized by considering different train-test split ratios, numbers of hidden layers and neurons, and activation functions. The proposed model's performance was evaluated using various metrics and was compared with other commonly used machine learning algorithms. According to the results, the proposed model outperformed traditional machine learning algorithms by achieving higher accuracy and lower error rates, with the best-performing model having four hidden layers with 70 neurons per layer and a scaled exponential linear unit activation function, resulting in a coefficient of determination value of 0.89. This demonstrates the potential of deep neural network models for predicting truck productivity in open-pit mine sites. Moreover, a single variable sensitivity analysis was conducted to investigate the impact of input variables on truck productivity. The results show that haul distance is the most influential variable for the prediction of truck productivity.