978-1-7281-2530-5/19/$31.00 ©2019 IEEE Fault Diagnosis in Vehicular Networks Using Do-Calculus Baisakhi Chatterjee Department of Computer Science and Engineering Institute of Engineering and Management, Kolkata, India baisakhi.chatterjee95@gmail.com Rohan Laha Department of Electrical & Electronics Engineering Institute of Engineering and Management, Kolkata, India rohanlaha5@gmail.com Preetarpan Banerjee Department of Electronics & Communication Engineering University of Engineering and Management, Kolkata, India preetarpanbanerjee@gmail.com Phillip G. Bradford Department of Computer Science University of Connecticut, Stamford, USA phillip.bradford@uconn.edu Abstract: Real time fault diagnosis is an integral part of modern automobile industries. Vehicles are no longer a sum of their mechanical parts but are built up of a number of interconnected programmable components. ECUs (Electronic Control Units) drive these components and that means if one component fails, it may lead to a complete failure. Research has thus focused on effectively solving this problem in real time. In this paper, we aim to present a working model that using Bayesian networks and intervenal calculus that accurately diagnoses faults in vehicular systems. Keywords: Fault, Fault diagnosis, IoT, Internet of Things, Bayesian network, Do-calculus. . I. INTRODUCTION The advent of internet computing and boom of internet of things have brought the need for fast Swapnil Mitra Department of Electrical & Electronics Engineering Institute of Engineering and Management, Kolkata, India swapnil.mitraofficial@gmail.com Dattatreyo Roy Department of Electrical & Electronics Engineering Institute of Engineering and Management, Kolkata, India dattatreyoroy@gmail.com Himadri Nath Saha Department of Computer Science and Engineering Surendranath Evening College, Kolkata, India contactathimadri@gmail.com analysis and processing of data in order to configure systems for maximum efficiency. Internet of Things [1] is a catch-all term that refers to the multitude of devices that use the internet to connect, process and exchange data as required. Embedded computer systems make it to program this information gathered. Thus, vital that we utilize the data in order to develop better functionality. Fault detection in computer systems is one such area of research. Real-time fault diagnosis is an immensely interesting area of study. In this paper, we focus on fault detection in vehicular systems. Vehicles in modern times are no longer a sum of their mechanical parts but are complex and intertwined programmable components. ECUs (Electronic Control Units) are the basic components of automobile control systems. Sometimes, if one component fails, a complete failure may occur. Thus, it is essential that such faults are found and fixed as soon as possible in order to curb vehicular failure. Researchers are trying to formulate better and better models in order to accurately diagnose and repair 1043