Combined Application of CRITIC and Back Propagation Neural Network Prediction Method for Vehicle Emission Parametric Evaluation in Logistic Networks and Distribution Systems

Sunday Ayoola Oke (Department of Mechanical Engineering, University of Lagos, Lagos, Nigeria)
Alexander Iwodi Agada (Department of Mechanical Engineering, University of Lagos, Lagos, Nigeria)
Bayo Yemisi Ogunmola (Department of Mechanical Engineering, University of Lagos, Lagos, Nigeria)
Henry Ogbemudia Omoregbee (Department of Mechanical Engineering, University of Lagos, Lagos, Nigeria)
Modupe Adeoye Onitiri (Department of Mechanical Engineering, University of Lagos, Lagos, Nigeria)
Nehemiah Sabinus Alozie (Department of Mechanical Engineering, University of Lagos, Lagos, Nigeria)
John Rajan (School of Mechanical Engineering, Vellore Institute of Technology, Chennai, India)
Swaminathan Jose (School of Mechanical Engineering, Vellore Institute of Technology, Vellore, India)
Pandiaraj Benrajesh (Dexterity Design Services, Chennai, India)

International Journal of Industrial Engineering and Engineering Management
Vol. 8 No. 1 (2026)

Logistic network and distribution services in the packing industry generate vast amounts of vehicle emissions, which are difficult to control due to the lack of scientific guidance. This paper predicts the emissions from vehicles engaged in logistic networks. The CRITIC (Criteria Importance Through Inter-criteria Correlation) method is integrated with the backpropagation neural network (BPNN) to analyze the input parameters of the emissions process from vehicles plying roads in the packing industry. The packing industry transports goods in packages and delivers them to customers in a road network. The application of the integrated CRITIC-BPNN method is demonstrated by using the dataset of a packing industry in India, obtained from the literature. The results revealed how the packing industry can translate its environmental control strategy into a parametric framework, yielding diverse outputs to assist in selecting optimal decisions. The proposed method exhibits unique characteristics. (1) It emerges as the first decision-making tool for objective criteria-based vehicle emission process control and could serve as a tool for logistics planning. (2) It employs integrated ideas of multi-criteria analysis and neural networks. (3) Objective and optimal emission analysis decisions are advanced through an analysis of multi-criteria tools. By employing the CRITIC method, the conflicting objectives of the emission process are successfully tackled and synchronized objectively to arrive at robust weights used in the back propagation neural network method for the final prediction. Moreover, the neural network method uses intelligence to gather data and translate it into usable forms to predict vehicle emissions.