www.astesj.com 287 Optimization of the Procedures for Checking the Functionality of the Greek Railways: Data Mining and Machine Learning Approach to Predict Passenger Train Immobilization Ilias Kalathas * , Michail Papoutsidakis, Chistos Drosos Department of Industrial Design and Production Engineering, University of West Attica, Athens, 15354, Greece A R T I C L EI N F O A B S T R A C T Article history: Received: 17 June, 2020 Accepted: 19 July, 2020 Online: 28 July, 2020 Information is the driving force of businesses because it can ensure the ability of knowledge and prediction. The railway industry produces a huge ume of data, with the proper processing of them and the use of innovative technology, there is the possibility of beneficial information to be provided which constitute the deciding factor for the correct decision making. Safety is the railway comparative advantage that has to be reinforced by each business administration while making the optimum decisions. The main purpose of this paper is the investigation of the most important dysfunctions that arise in a train and can cause its immobilization at the main passenger rail, resulting in huge delays of conducting the routes setting the passengers at risk. Afterwards the total of malfunctions is assessed and the most important, potentially, malfunction is assessed, so as the executives of the Greek Railway company to plan and redefine the processes and the initial plan of the predictive maintenance. This paper demonstrates the effort of implementing innovative applications by making use of methods from the rapidly developed field of Data Mining to the Greek Railway Company that uses obsolete procedures for the control of the trains’ functionality in order to investigate the data for the provision of specialized information which will be used as a tool for the faster, more accurate and precise decision making. This decision making approach is based on a specific algorithm’s design in order to automatically detect faults and make periodic maintenance of trains easier. Holistic approach is performed in the management of real data from the Greek railway industry and a predictive model of Machine Learning is developed, for the optimization of the management’s performance of the trains reinforcing the strategic target of the railway industry which is the transportation of citizens with safety and comfort. Keywords: Machine learning Railway Train Immobilization Data mining Predictive Model Malfunctions Diagnosis 1. Introduction Businesses are more and more using sets of data in order to conduct their decisions. Developments in the field of Data Mining and the Machine Learning are expected to predominate in 2020 and to create, within the next decade, significant opportunities to all the companies [1]. The emerging technologies change the way that businesses collect and extract useful information from the data [2]. The Data Mining is an effective method for the analysis of a huge amount of collected data that extracts useful information. The Machine Learning is a field that was developed by the artificial intelligence and assists planning and development of algorithms and the eution of the performance related to empirical-operational data [3]. The railway industries are a field the performance of which progressively depends on their ability to extract information from complex sets of data and take the optimum action in real time [4]. The Data mining in conjunction with the Machine Learning has the capacity to improve the operational progress raising the level of efficiency in decision making and the overall procedures [5]. The railway is a system of passengers’ and merchandise transportation with wheeled vehicles (trains) that roll on rails. The railways as a mean of transport is defined by three components, the functioning utilization, the infrastructure, and the rolling stock. With the term rolling stock we refer to all kinds of vehicles pulled or driven on rails that perform railway transports [6]. The Railway Rolling Stock is a complicated system the smooth operation of which plays a leading role on the exploitation of the rail system and is subject to progressive lesions (wears, erosions, malfunctions ASTESJ ISSN: 2415-6698 * Corresponding Author: Ilias Kalathas, University of West Attica, Tel: +306974731434 , i.kalathas@uniwa.gr Advances in Science, Technology and Engineering Systems Journal Vol. 5, No 4, 287-295 (2020) www.astesj.com https://dx.doi.org/10.25046/aj050435