Traditional software development's effects on safety

Abstract

Faults in a system can result in catastrophic consequences such as death, injury or environmental harm. For example, the Therac 25 incident killed two patients and severely injured a third patient due to a software error. The Therac 25 is a computer-controlled therapeutic radiation machine. Governmental, industrial, and academic researchers are searching for new ways to prevent and detect hazardous faults when developing and certifying safety-critical software systems. To determine the effects of these new techniques, this paper discusses an experiment where developers in the control group followed a traditional methodology while the experimental group followed a modified methodology, which consisted of the control group's methodology along with certain safety-specific methods and guidelines for the following life-cycle phases: specification, design, and verification. The results showed that the experimental group had fewer latent safety-critical faults than the control group.<>