Safety and Security Assurance for Engineering High Confidence Software in Information Systems

点击翻译按钮获取中文摘要

This discovery grant application is a continuation of my ongoing research program on security assurance in complex information system software, including an extended focus on safety. Software-based information systems (IS) are playing an increasingly critical role in knowledge-intensive industries, e.g., healthcare, defense, transport, energy, etc. Despite anticipated benefits of deploying and networking such systems, there has been mounting evidence on risks and hazards. Safety, Security and Privacy concerns are at the forefront of these concerns. Example hazards (in healthcare) include misdiagnosis and wrong treatment decisions and confidentiality breaches that may lead to identity theft, loss of reputation and livelihood.Given numerous cases were IS software faults caused harm or suffering, regulators and users have called for better software engineering practices. Comparisons are often made with more traditional industries that employ critical software for automation and control (e.g., manufacturing, transportation, and power generation) and have a higher degree of maturity when it comes to quality assurance.However, IS software is significantly different from automation and control software and the models, methods and tools established in these more traditional industries cannot simply be transferred to knowledge-intensive industries. In the light of these difficulties, regulators have shied away from a product-focussed approach to quality assurance and adopted a process-focussed strategy instead. In other words, rather than assessing direct evidence about safety and security of the IS software products, regulators focus on the software development processes employed in developing the product (indirect evidence). This approach has been criticized as ineffective by many researchers, e.g., Parnas: "A rational design process: How and why to fake it", Wassyng et al. "On Software Certification: We Need Product-Focused Approaches". The development of product-focussed software assurance methods is seen as an important research challenge and a prerequisite for enabling truly evidence-based software certification practices. The objective of my research program is to research and develop innovative models, methods and tools for generating product-focussed (direct, evidence-based) quality assurances for the safety and security of critical IS software. Specific objectives include the development of(1) a system-theoretic IS hazard model (SIHM) as a foundation to analyze safety & security concerns and discern software-induced incidents from environmental factors, e.g., human error;(2) methods to engineer high confidence IS software (and systems of systems) based on the foundational SIHM and conducive to product-focussed assurance; and(3) product-focussed methods to verify, validate and certify safety and security-related properties of high confidence IS software.Our scientific approach involves (1) theory-building using a mixed method approach, (2) reduced-scale prototyping as exploratory case studies and proofs of concepts , and (3) empirical studies of applying the proposed methods "at scale" to real-world IS software. We will primarily study healthcare IS but also other domains, e.g., defense, aviation.The proposed research will generate novel and high impact results that will enable software engineers and regulators (e.g., Health Canada) to implement safer IS software. The results will help save lives and avert harm from Canadians. Student trainees will play an important role. Students at all levels will be trained in a highly collaborative, team-based setting, with exposure to other disciplines as well industrial and academic collaborators. HQP with software assurance skills are in high demand in Canada.