Neural model for risk estimation in transport service delivery

Abstract: The article presents a concept for estimating risk in transport service delivery using artificial neural networks. ANNs are mathematical models that perform information processing modeled on the functioning of neurons in the human brain. The presented neural model is based on real-world results of risk estimation in transport services.The disadvantages of this method include the need to collect a relatively large amount of data and the labor-intensive nature of the network building stage. Despite this limitation, the use of ANNs in practice eliminates random errors and enables obtaining a reliable risk assessment necessary for making decisions regarding the implementation of a transport service.
1. INTRODUCTION
The dynamics, variability, and instability of the socio-political situation can drastically affect the quality of a transport service (TS). Awareness of existing and potential risk values in service delivery provides greater freedom of action toward specific solutions. The greatest risk is not analyzing risk.
Therefore, the ability to accurately determine all conditions and aspects affecting the proper, effective, and efficient implementation of TS becomes of significant importance. This requires the development of clear principles and methods for estimating and managing risk. Their application will enable the elimination or at least the reduction of its value to an acceptable level. Risk assessment should be a continuous process conducted at all stages of TS implementation. The most important and difficult element of the assessment, which influences subsequent actions, is risk estimation.
Identifying potential threats as well as their sources and causes requires extensive knowledge and experience in technical, logistical, organizational, and marketing issues. The classification of risk in terms of acceptability criteria, the determination of priorities, and the engagement of appropriate resources will depend on the probability values of the occurrence of threats and their consequences. The results of this assessment will determine the decision to implement the TS, and consequently, the implementation of planned actions leading to the reduction of the risk value to an acceptable level. The costs of potential losses incurred as a result of an improperly implemented TS are a derivative of the corrective actions taken. (…)

Neural model for risk estimation in transport service delivery – PDF file for download