M-Zones - - Workpackages

M-Zones Workpackages

The philosophy of this work package is to build a coherent team between the three institutions, maximising the synergies available within the programme while allowing partners to focus on their core areas of expertise.
The workpackages will produce an annual report that will be available to the HEA and its nominated experts for review and evaluation. The annual report will incorporate information on the progress made to the realisation of the research programme objectives during the reporting period. It will also address the key drivers and collaborations developed throughout the reporting period. It will conclude by providing an overview of its work plan for the next period, including information on any major technical developments that can impact on its programme goals.

This work package will provide an initial review of the state of the art in key areas, and ensure that this information is kept up-to-date on a rolling basis. The intention is to ensure that the programme's work is not "blindsided" by industrial developments or emerging standards. By maintaining a rolling surveillance activity the programme will avoid becoming "fossilised" at an early stage of device evolution - unacceptable in a fast-moving area.

This work package will develop a suitable abstract Management Architecture for design and implementation of the interconnection of Management Smart-space Zones (or M-Zones).
Real device characteristics will be collected from the WP4 test-beds and used to provide a "library" of characterised devices. An additional set of "possible" devices characterisations will be developed to model potential future directions in technology - for example portable video streaming clients or instant-messaging products. Another interesting category of "possible" device is the management device: a smart artefact whose sole purpose is to provide management information or functions within a smart space.
The interactions of the devices will also be investigated so develop and refine the models underlying the simulation and trial work being undertaken in WP4.
This workpackage will define an initial M-Zone management architecture in year one and continue to refine and update this architecture throughout the work programme lifetime. The architecture is seen as a key deliverable of the work programme.

The programme is founded on the notion that much of the previous and current work in smart spaces has focused exclusively on the creation of small, single smart spaces. A large amount of individual analysis and design is performed to build such systems - they are crafted rather than being engineered - and their final properties are unclear during the design phase. Improving this situation, in order to allow organisations accurately to predict the behaviours and limitations of systems and their evolution, is the core target for the programme.
This work package will examine a number of possible approaches to this problem, drawing the technical results in the other work packages into a coherent engineering framework. Among the approaches we consider potentially valuable are a set of guidelines for systems designers contemplating new smart spaces, and a "smart space CASE tool" allowing computer-aided software engineering of smart spaces by allowing the implications of alternative design decisions to be modelled and simulated.

The goal of this work package is to construct and maintain an evolving set of test-bed environments, one per partner, within which to evaluate the tools and technologies developed. Although expensive, such test-beds will allow the programme to address the real-world engineering aspects of its research as well as the theoretical underpinnings.
Complementary to the test-beds, the simulation work package will provide a purely virtual testing environment for investigating technologies. The advantages and disadvantages of simulation are that it allows experimentation with far larger and more complex systems than can be constructed by a small project, can explore the implications of yet-to-be-realised developments, but does not necessarily give insight into the engineering aspects of deployment. For this reason a combined testing and simulation framework is to be preferred.

The final work package of the programme will collect together the evaluation aspects of the entire programme. This is preferable to individual, per-package evaluation tasks as it allows the results of experiments to be analysed against a complete set of deliverable technologies.
Evaluation will be based around a set of scenarios - "stories" describing possible interactions with a smart space. Each scenario will be explored using the different techniques under investigation: modelling, simulation and (where possible) implementation.
This workpackage is also responsible for the dissemination of the project results through conferences, workshops and journals as well as through submissions to key standardisation work. In addition an interactive website will be developed to support interaction between the project and other internally renowned research teams within the area of smart space management. By its nature the identification of the dissemination channels is well be carried out in real time. At the end of each year a report will be produced in the form of a deliverable indicating the dissemination activities carried out within the previous 12 months.

Adaptive Policies form the core of the management framework being developed within the programme. Policy-based management allows system engineers and administrators to state the functions they wish to allow (or disallow) in a system, the properties they wish to maintain, and so forth. The key is that this expression occurs at the level of the system's observable behaviour rather than in terms of implementation mechanisms, allowing the management system to deploy a range of low-level approaches to realise the high-level policy goals.
Work on policies will develop a policy capture and expression framework based on extensive previous work by the partners and others. Policy expression must be sufficiently generic to allow expansion but sufficiently specific to permit implementation, since a policy which is stated but not monitored or enforced is not useful. This is a difficult task, which will be helped by the partners' extensive previous experience in the area.
The second thread is to provide mappings between the inputs and outputs implied by a policy system and the three strands of management under investigation: WP2/3's abstract model (providing extra constraints to the model), WP5's simulations (allowing the effects of policies to be examined in a range of situations), and WP5's test-beds (providing vital real-world engineering information). This breadth provides the widest possible range of analyses into different policy approaches and the assessment of their impacts.