STRATA; Layers for Structuring Trustworthy Ambient Systems

For computing as a whole, and ambient systems in particular, significant new challenges will have to be faced because of the rapid evolution in Cyber-Physical Systems (CPS). A major impetus towards CPS is the promised Internet of Things (IoT). The scale of the IoT could soon involve trillions of devices generating masses of real-time data and thus demanding unprecedented power. To sustain this scaling and huge surge in energy demand, ambient systems will need to be designed to trade off power, performance and reliability. Inevitably, this will require ever more sophisticated responses to achieve trustworthiness.

Our TrAmS-2 platform was shaped by two important factors affecting ambient system design. First, the power provision/consumption of devices, rather than cost, was becoming the limiting factor in the deployment of ambient systems. Second, novel paradigms such as cloud computing offered a new dimension of ambience in that data and programs can be migrated without physical movement of agents.

The TrAmS-2 platform grant has sustained our research group which has created new projects on sound technical foundations, methods and tools to model, design and analyse Trustworthy Ambient Systems. Our team, and the projects in which it is involved, are developing methods for designing mobile devices with energy-efficient and power-constrained hardware. Our project portfolio includes 13 EPSRC, EU, industry and other projects with applications in sectors including automotive, rail, space, business, healthcare and consumer electronics. Our current arsenal of tools and methods includes the powerful toolsets of Rodin, Workcraft and Symphony, advanced patterns for modelling fault tolerance and world-leading techniques in proof technology, simulation and ample evidence to support industrial deployment of formal engineering methods. With such a solid foundation we will enter Strata well equipped with formal engineering methods, advanced tool support, architectural and algorithmic approaches for embedded systems design and modelling, in particular capturing systems with multiple modes targeted at power and reliability.

In Strata, the research platform will receive extra impetus through a two-pronged attack on the challenges of future ambient systems design with the teams in Newcastle and York working together. This will facilitate making a qualitative step in terms of rigorous and model-based approaches to the design and analysis of future resource-limited ambient systems. The cornerstone of the platform approach and methodology is a shared vision that future complex ambient systems have to be structured in layers. These layers, or strata, of system resources and delivered functionality in terms of time and power will be combined with cross-layer fault tolerance and even adherence to the (levels of) specification in an environment where one has to accept that components will fail. Strata will address the challenges in four interlinked themes: methodology, cross-layer fault tolerance, real-time layering and real power.

The skills of the three involved research groups are complementary and form a solid research base in software, systems and microelectronics to establish a unique team in terms of capability and expertise, with international profiles in formal methods, dependability, real-time and energy-modulated systems. Strata will provide continuity for research staff, encouraging new, risky, research in areas created by this novel mix of expertise.

Strata's way of generating impact, both economic and societal, fits with our scientific approach in that it can be understood as bands of impact activity. One band relates to the existence of the platform grant itself and the work that will be carried out within it: we will have an impact on the people involved in it, on the academic community and on the wider society. The other band of impact will come from Strata offspring projects because of the new avenues of research that will be explored in structuring methods for complex ambient systems with limited resources.

Within the first band, we envisage producing strong societal and economic impact of our research through the effect of the new methodology on the practices of developing complex fault-tolerant mixed-criticality systems with limited resources through the radically new concepts of layers in systems structuring.

The extension of TrAmS-2 into real-time and layered ambience offers great potential for impact in broad areas of industry and society. Beneficiaries include businesses providing infrastructure, devices and services for ambient systems, and end users for whom major concerns are the security, safety and quality of experience provided by the system as a whole.

Several groups have a stake in ensuring fault-tolerance and resilience of ambient systems. Economic beneficiaries include companies and sectors that develop software and services for such systems, for whom the cost of development, modification and upgrading directly depends on the complexity of the systems. The disciplined application of the cross-layer fault tolerance will allow them to reduce the cost of development and make their products more competitive.

The entire ICT sector is being strongly influenced by societal and economic demands for systems to be green, energy-frugal and power-proportional. Through the promotion of better understanding of the holistic nature of system design issues, Strata will promote methods and tools for rigorous development through the interplay of power and timing constraints. This vision will be passed to existing and new SMEs and R&D organisations, particularly in the embedded systems sector, where the drive for smart-energy utilisation is increasing.

In the second band we will ensure that Strata enables the impact of the new projects that result from the platform grant. Our approach is therefore to ensure that the Strata group interacts with potential beneficiaries of these "offspring projects", and takes account of their needs when formulating research strategy and developing project proposals.

A major impact route is through developing tools supporting the well-founded modelling and analysis methods for dependable ambient systems that Strata offspring projects will generate. We will build on our extensive experience at developing usable tools that have achieved industry deployment. To overcome the difficulty of accessing formal methods tools, we will set up the infrastructure to allow new and existing tools to be wrapped with standardised service-based interoperable interfaces. This approach is intended to increase, by orders of magnitude, the number and range of organisations that can then benefit from these tools. In particular it will allow SMEs to benefit from tooling that has hitherto been limited to large organisations with expertise and computing power.

All the investigators have established records of creating impact from their research both from the previous TrAmS-2 Platform Grant and other extensive research careers. Much of the existing evidence is of solid industrial impact via projects that benefited from the previous tranche of TrAmS-2 funding.