Beside PhD research projects, AICAUSE research is organised in projects either funded by industry partners and/or competitive grants.
(Both so-called specific research projects in the Research Framework Agreement) or pre-studies aiming at grant proposals targeting government or industry funding.
Current and future specific grant and PhD projects are described in the following broad research program themes:
- Parallel and Distributed Platforms Architecture Design
- Cyber-physical Systems Verification and Testing
- Virtual Collaboration, Platforms and Agility
Parallel and Distributed Platforms Architecture Design (PADx2)
CPS are widespread in safety-critical domains such as vehicles, production machinery, aircraft or medical devices. Failures of these systems may lead to considerable loss of money or even endanger human lives. This is one of the most challenging areas of system engineering because of the complexity of computing and physical dynamics, especially in distributed control of interconnected systems. The main challenge is to combine two worlds - the physical and the virtual one – to ensure that software components operate in discrete program steps. Meanwhile, the physical components evolve over time intervals in line with physical constraints. Control is increasingly distributed and decentralised requiring the coordination of systems with physical and virtual resources spread across devices and across sites. CPS architecture is concerned with structure and design of these kinds of systems and to develop and improve architecture-centric development processes, which can be adapted and scaled to specific needs of projects.
Areas of specific interest include: software product lines, reuse and evolution roadmaps for self-adapting or fault-tolerant component and service-based architectures, with emphasis on deploying new components in running service-oriented ecosystems.
Cyber-Physical System Verification and Testing (CPSVT)
Current activities include spatial aspects of cyberphysical systems (CPS) and cloud-enabled statistical testing of CPS.
Guaranteeing properties of software systems by using mathematically founded reasoning (software verification), such as showing the absence of errors in program code, is a well-established research topic. Different techniques and tools such as model checkers and theorem provers have been established to support and automate verification tasks. In our work, we are adapting existing software verification techniques and establishing new methods to explore and verify properties of the spatial behaviour of automation systems that are controlled by software, including exploration of collision freedom, safety and spatial aspects of quantitative performance characteristics.
The focus on efficient and effective distributed and remote testing for cyber-physical automation systems (robotics) over clouds is associated with:
- novel testing methodologies for the cost-effective test case selection and light-weight test result verification for parallel execution from the cloud;
- enabling testing as a service over the cloud with the aim of improving the testing performance and reducing its cost; and
- the development of cloud-enabled testing reference model to inform CPS design for testing and to increase testability of deployed systems by design.
Research interests in this program include software safety and reliability monitoring for robotic or automated physical systems directly co-operating with humans, operating on humans, or in closest proximity around humans with risks of injury or endangering human lives. The research also focuses on spatio-temporal aspects of model-checking and model-driven testing in addition to novel cloud-enabled statistical testing methods. Furthermore, it looks at cost-effective test case selection and light-weight test result verification for widely distributed systems as well as high-efficacy testing-as-a-service in the cloud for cyber-infrastructures and the architectural design for testability.
Virtual Collaboration, Information Architecture and Agility (VCI2A)
Research on software architecture and its intertwining with multi-domain development processes is central to this research program. Particularly in relation to quality and evolution designed to reduce life cycle costs and increase quality (maintainability, performance, reliability) of similar large scale/legacy automation software systems. Product line architecture is the lingua franca and reference for the views of different stakeholders and formal domain-specific documentation and artefacts from different phases, projects and parties that come together to deliver the overall software-intensive automation project.
Modern software platforms supporting such development processes are heterogeneous by nature, and address evolution, architecture recovery, refactoring, planning, quality, cost analysis, beside the traditional lifecycle phases from analysis through design to implementation. Empirical case studies include refactoring automation design tools or refactoring automation systems to service-oriented architectures, parallel and distributed architecture and system verification and testing for safety.
Points of interest for this research program include empirical global software engineering research, cloud-enabled virtualisation of laboratories and collaboration, tele-presence and visualisation infrastructures. It also develops agile teams and methods for rapid software-enabled innovation and productivity gains, with larger numbers of smaller teams operating across time zones, and with very short incremental design-build-and-deploy cycles (measured in hours or days rather than months or years).