Safety-critical software systems are increasingly affecting our lives and welfare as more and more software is embedded into cars, medical devices and airplanes each day. New approaches and international standards are being developed to ensure the safety of these systems before they are delivered.
Safety-critical domains are characterized by heavy regulatory demands to which companies must adhere to before they can place their devices on the market. Regulatory audits are conducted regularly to evaluate these companies and the safety of their systems and devices. In order to pass these audits, system developers have to ensure that all regulatory requirements have been adhered to in the design and development of each of the safety-critical subsystems. Software development companies moving into the safety-critical domains often find themselves overwhelmed by the number of regulatory requirements they need to satisfy before they can market their systems or devices.
Safety-critical domains such as automotive, healthcare and medical devices, space and nuclear energy all have a number of common challenges when software is being developed and embedded into their systems. Namely, adopting a safety- and risk-centric approach to the software and system development while facing the very same commercial demands of speed and access to market as generic software developers. How then can agile software development methods be used efficiently in safety-critical domains while meeting the safety, interoperability and security demands of safety-critical systems? What are the additional requirements for engineering processes in safety-critical domains and how could they be met? How can we ensure that software in safety-critical systems is interoperable with the rest of the system and without any internal or external security vulnerabilities?
track focuses on the issues of software safety, interoperability, processes and
process models, and agile methods in safety-critical domains such as
automotive, health and medical devices, and nuclear. There are altogether five
paper presented in the Safety-Critical Systems track at QUATIC 2016. Following
is a short description of each of the papers.
The first paper of the track discusses process assessments in the nuclear energy domain as an input to a safety assurance case that increases the safety of the software systems. The authors illustrate how assurance needs can be met with an exemplar process assessment method in the nuclear energy domain. The main result of the paper is an analysis of the assurance case and how a process assessment method together with the assessment results can be used as evidence in regulatory safety assurance.
The second paper presents a view of agile software development and quality assurance in complex safety-critical systems. Although agile software development methods offer safety-critical system developers more advantages for problem discovery, their design review methods rely on validation by the development team. In order to reduce software errors in safety-critical systems, these agile teams should consider how software and system designs might be validated by external teams instead.
The third paper explores software quality in automotive systems. The research presented is motivated by supporting the time-consuming safety-case development by automating it. The authors propose to automate the documentation generation in compliance with the standard ISO 26262 (functional safety in automotive domain) and to semi-automatically generate the required safety case.
The fourth paper brings us back to the medical device domain where safety assurance cases are not yet widely used. The authors argue that the number of medical device recalls due to software failures could decline when process compliance is supported by safety cases. In addition to that, the authors explain how safety of medical device software systems may further be improved when the industry could learn from the previous failures across the medical device industry.
Finally, there is a paper describing the use and
testing of mobile robots in healthcare environments. With the rise of mobile
robots in common households, these robots could also be used in different
safety-critical domains and environments. The authors describe the testing of a
mobile robot in a hospital environment to develop a generic acceptance test strategy for mobile robots’ navigation algorithm.
Dr. Marion Lepmets is a Senior Research Fellow at the Regulated Software Research Centre in Dundalk Institute of Technology, Ireland. Her research focuses on the software development processes and their compliance to regulatory requirements in the medical device software domain. She has co-authored the Process Reference Model for IEC 62304 (IEC TR 80002-3), which was published in 2014. Her postdoctoral research, funded by the National Research Fund of Luxembourg, addressed process improvement impact on IT service quality. She has conducted research in process improvement and process assessment since 2000, graduated from Tampere University of Technology (Finland) with Dr. of Technology in 2007, and has been teaching process management courses in both Tallinn University of Technology and Tartu University in Estonia. She has been involved in the development of software engineering standards in the International Standardization Organization (ISO/IEC JTC1 SC7) for the last 10 years as a national delegate of Estonia, Luxembourg and Ireland. For the last three years, she has also been involved in the IEC SC62A JWG3 working group that is responsible for development and maintenance of the International standard for medical device software lifecycle processes (IEC 62304).
Thematic Tracks >