NESSI has published a paper on New Directions for Software R&I in the context of future European collaborative programmes.
Recent developments in AI will have significant consequences in many software-intensive application areas, although much remains uncertain. Traditional approaches may prove insufficient to deliver safe and dependable systems.
Future R&I in software needs to address several priority topics, with a particular focus on critical infrastructure and systems:
- Management of software systems as integral parts of systems in and across different domains which directly affect people (e.g. healthcare, energy, manufacturing, transport).
- AI-native software engineering techniques and tools to ensure software is robust, secure, safe and fit for purpose, and addresses the full lifecycle of systems.
- Engineering methods and tools that assure (and evidence) secure-by-design, safe, compliant and privacy-preserving operation in many and varied environments, each with their own stakeholders, purposes, constraints, regulations, priorities and risks. Facilities for appropriate and effective human oversight and accountability are essential.
- Development of libraries and frameworks focused on significantly improving resource efficiency (computing hardware, networks, energy, water) across heterogeneous computing environments
- End-to-end security facilities (including quantum-resistant and post-quantum cryptography) to support confidential and privacy-preserving computing on untrusted edge and cloud systems.
- Ensuring that human skills are maintained and developed so that effective application of AI technologies is broadly accessible, including to SMEs. Effective education and training of current and future practitioners is essential to achieve and maintain sovereignty in the face of the anticipated radical changes in software and services.