Process Mining of Mining Processes: Analyzing Longwall Coal Excavation Using Event Data

IEEE Transactions on Systems Man and Cybernetics Systems, vol. 54, pp. 2723–2734

Abstract

The mining industry faces many challenges, prompting the adoption of new technologies and continuous improvement of processes to improve operational efficiency and personnel safety. Using data from information systems combined with novel process mining (PM) techniques creates new possibilities for improving industrial processes. This article presents a comprehensive method of modeling and analyzing the longwall process in underground mining based on event data using process mining (PM4LMP). The method comprises four basic steps: 1) data gathering; 2) data preprocessing; 3) creation of event logs; and 4) PM tasks. In our method, we proposed, among all, case ID identification based on heuristics using context data and activity identification with supervised and unsupervised approaches, which provide complementary information about process execution. The method assumes an in-depth analysis of processes based on sensor data and knowledge gathered in IT systems, which can significantly improve the quality of information at managers’ disposal when making decisions regarding the mining process.

Authors 3

1. Edyta Brzychczy

   AGH University of Krakow

   Affiliation as printed

   Faculty of Mechanical Engineering and Robotics, AGH University of Kraków, Kraków, Poland

2. Agnieszka Żuber

   AGH University of Krakow

   Affiliation as printed

   Faculty of Civil Engineering and Resource Management, AGH University of Kraków, Kraków, Poland

3. Wil M. P. van der Aalst Aachen Process and Data Science Group

   RWTH Aachen University

   Affiliation as printed

   Process and Data Science Group, RWTH Aachen University, Aachen, Germany