Enhancing Reliability in Heavy Duty Autonomous Mobile Machines through Fault Tolerant Edge Computing

Abstract

This paper presents a novel fault-tolerant edge computing architecture for heavy-duty autonomous mobile machines in industrial environments. The proposed system integrates two virtual machine hosts with a circular topology of four Ethernet switches, ensuring network re-silience and operational continuity. A key feature is the implementation of automatic takeover protocols, enabling seamless transition between hosts during hardware failures. The network infrastructure leverages Rapid Spanning Tree Protocol (RSTP) and additional loop protection mech-anisms to maintain stability and achieve rapid fault recovery. Virtual machines running Ubuntu Linux or FreeBSD are strategically deployed to handle specific tasks with critical services replicated for enhanced re-liability. The system incorporates advanced data management through a PostgreSQL database with master-slave replication. Comprehensive fault tolerance mechanisms, including redundant connections and grace-ful degradation capabilities, ensure robust performance in challenging industrial settings. This architecture significantly enhances the reliabil-ity and autonomy of heavy-duty mobile machines, addressing the critical need for uninterrupted operation in industrial automation applications.