This project involves safety and fault-tolerance problems raised in the development of safety-critical computing systems. There are many computer applications in which safety and not reliability or availability is the overriding concern. Reduced or altered functionality of such systems is acceptable as long as no harm is done. Theoretical research will focus on the specification and design phases of the software lifecycle. Specifically, the project will develop a framework of definitions of safety, establish the responsibility of the software engineer within that framework, and design software architectures to support the development of safe software. It will investigate the possibility that an enhanced form of software reuse can be exploited to facilitate rapid development of specifications and designs for safe software. Empirical research will focus on a case study which will involve certain aspects of the software for an experimental safety-critical medical device. The device, known as the Magnetic Stereotaxis System, is a new approach to various forms of therapy for brain disorders. Software safety specifications will be constructed and a prototype built. Reuse libraries populated by a small number of parts derived from the case study will be built and used in subsequent studies of reuse technology.
