International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 185 - Number 29 |
Year of Publication: 2023 |
Authors: Sorle Bumaa, Onate Egerton Taylor, Nuka Dumle Nwiabu |
10.5120/ijca2023923049 |
Sorle Bumaa, Onate Egerton Taylor, Nuka Dumle Nwiabu . Faults Detection and Isolation System for Internet-of-Things Device (Autonomous Vehicle). International Journal of Computer Applications. 185, 29 ( Aug 2023), 55-59. DOI=10.5120/ijca2023923049
This study proposed an automated system for faults detection and isolation of internet-of-things (IoT) device with special emphasis to an autonomous vehicle (i.e. self-driven car). Fault detection can be described as the process of identifying faults in physical systems in order to proffer solution to the discovered faults. The developed system was achieved using Structured System Analysis and Design Method (SSADM), Hypertext Pre-processor (PHP) and MySQL. In addition, the developed system was specifically designed to address sudden abnormal faults such as steering and brake stiffness in the autonomous vehicle. The proposed system was developed using Server-alert data synchronization technique which enhances the system to stream the system event log file in order to detect and isolate faults when the system is on motion. The proposed system also provided an emergency backup framework for the stiffed steering and brake faults. The system was tested and evaluated using parameters such as the number of iterations carried out, the number of inputted data for synchronization per iteration, the number of inputted sensor codes per iteration, the number of detected faults of the autonomous vehicle, the time in seconds taken to detect the fault and the system suggestion in addressing the detected faults. A total number of 5 iterations were carried out each for the existing and proposed systems respectively. The proposed system used a speed of 12 seconds to detect and isolate faults of an autonomous vehicle while the existing system used 27 seconds. Furthermore, the system was also tested and evaluated using confusion matrix evaluation technique. Parameters for the adopted confusion matrix evaluation technique encompasses true positive (TP), false positive (FP), true negative (TN) and false negative (FN). The developed system achieved a TP value of 89, an FP value of 5, a TN value of 3 and an FN value of 3. These values were further evaluated using the total number of correct possibilities all over the total number of possibilities made. Hence, a performance accuracy rate of 92% was obtained for the developed faults detection and isolation system.