A simplified overview of handling faults in systems around us

Abstract

Faults cause the observed behavior of a system to deviate from the expected behavior. They occur throughout the lifetime of a typical system. They may occur in the design phase of the system, the manufacturing phase, or during normal operation. In this paper, we present a simplified overview of handling faults in electronic as well as nonelectronic systems such as systems in engineering, science, biology, etc. The paper first defines the concepts of faults, errors, and failures. It then demonstrates the high cost of failure via several examples. The goal of this paper is to present a detailed and simplified discussion of the methods of handling faults in systems around us including fault avoidance; fault dismissal; error detection; fault location; error correction; fault masking; fault tolerance; and reconfiguration. Faults in a system can be divided into three categories: Faults that can be avoided (at a substantial cost), faults that can be dismissed due to various reasons, and faults that must be handled correctly before they become errors and ultimately lead to the overall failure of the system. The paper discusses the various techniques that prevent the faults in the system from leading the system to failure. The paper also discusses the requirements for fault tolerance and the methods used to achieve the desired fault tolerant capabilities in a typical system. The paper also presents two case studies to illustrate the concepts described above. The first system is the personal computer and the second system is the human digestive system. The case studies demonstrate the significant similarities between handling faults in electronic and non-electronic systems.