Objectives
The DCSC Safety Engineering topic aims to improve the effectiveness and cost-effectiveness of developing and certifying safety related computing systems. This is achieved through the development of safety engineering processes incorporating safety and hazard analysis techniques. As new technologies emerge, support will be provided for emerging classes of systems through the development of new techniques and processes. The key research areas of this topic are:
To maximise benefits to the company, it is a requirement that the safety processes and techniques developed under this topic are successfully integrated with the system development process to ensure correlation between the chosen design approach and the adopted safety strategy.
Projects currently being undertaken in this topic include supporting the use of integrated modular systems (IMS) and object oriented (OO) systems through the development of safety analysis techniques and guidance on modular safety case construction. Both of these support the production of more robust, maintainable and reusable systems.
Object orientation is a software development methodology that decomposes a system into objects, rather than decomposing on a functional basis. This allows systems to be designed that are more easily maintained and provide increased opportunity for reuse. The use of OO in the safety critical domain requires that it be possible to reason about the safety of individual classes or components in the system. Work is ongoing in this topic to adapt existing techniques such that the required safety properties can be generated. It is then necessary to understand how these properties and requirements are affected by design changes and reuse.
Integrated Modular Systems (IMS) are open distributed computer systems that are being used in many of the next generation of aircraft. IMS offers many advantages over current systems such as the ability to reconfigure in the presence of failure, the ability to upgrade hardware components with minimal affect on applications, and the opportunity for software reuse. These features cannot be used without a supporting safety process and modular safety case. Ongoing work in this topic is looking at modular safety case development to support such systems.