Current software development processes are expensive, laborious and error prone. They achieve adaptivity at only a glacial pace, largely through enormous human effort, forcing highly skilled engineers to waste significant time adapting many tedious implementation details. Often, the resulting software is equally inflexible, forcing users to also rely on their innate human adaptivity to find "workarounds". As the letters of support from the DAASE industrial partners demonstrate, this creates a pressing need for greater automation and adaptivity. Suppose we automate large parts of the development process using computational search. Requirements engineering, project planning and testing now become unified into a single automated activity. As requirements change, the project plans and associated tests are adapted to best suit the changes. Now suppose we further embed this adaptivity within the software product itself. Smaller changes to the operating environment can now be handled automatically. Feedback from the operating environment to the development process will also speed adaption of both the software product and process to much larger changes that cannot be handled by such in-situ adaptation. This is the new approach to software engineering DAASE seeks to create. It places computational search at the heart of the processes and products it creates and embeds adaptivity into both. DAASE will also create an array of new processes, methods, techniques and tools for a new kind of software engineering, radically transforming the theory and practice of software engineering. DAASE will develop a hyper-heuristic approach to adaptive automation. A hyper-heuristic is a methodology for selecting or generating heuristics. Most heuristic methods in the literature operate on a search space of potential solutions to a particular problem. However, a hyper-heuristic operates on a search space of heuristics. We do not underestimate the challenges this research agenda poses. However, we believe we have the team, partners and programme plan that will achieve the ambitious aim. DAASE integrates two teams of researchers from the Operational Research and Search Based Software Engineering communities. Both groups of researchers are widely regarded as world leading, having pioneered the fields of Hyper-Heuristics and Search Based Software Engineering (SBSE); the two key fields that DAASE brings together.

With the present trend towards miniaturisation of devices and development of microscale processes, there has arisen the need for microscale heat transfer equipment. For instance, in microelectronic processors, heat transfer is becoming the limiting factor in relation to increasing performance. Convective heat transfer with phase change offers significant performance improvements over single phase cooling systems. However, in microscale channels the mechanisms controlling the phase change and flow distribution differ from those at the macroscale. This project comprises an experimental study to solve some of the problems associated with phase change in microchannels and also the numerical modelling of boiling/condensation theory to both aid in design of experiments and develop design procedures for heat exchange processes based on two phase flow in microchannels.Development of novel experimental techniques (eg microsensor for heat flux and temperature measurement, fluid diode) will enable characterisation of local parameters in both boiling and condensation and active control of the boiling process.State-of-the-art numerical techniques will supplement the experimental investigation, enabling results to be applied to other systems. This combination of experimentation using innovative microsensors and control systems, developed using advanced microfabrication facilities, with numerical modelling offers a unique opportunity for the development of design procedures for such microscale heat transfer systems.

There is a growing demand that future networks should be ubiquitous, pervasive and multimedia-capable, i.e. anyone should have broadband access at anytime and anywhere. This demands that the future network access should be broadband, high-speed, and with Quality of Service (QoS) support for various multimedia applications. This vision will not become true without the development of next-generation QoS-enabled broadband wireless access technologies given that the current access technologies (e.g. WiFi, GSM/GPRS/3G UMTS and DSL cable access) cannot sufficiently satisfy the above requirements. Orthogonal Frequency Division Multiple Access (OFDMA) technology has emerged as a most promising transmission technique candidate to be utilized for the next-generation broadband wireless access networks. Compared to conventional single-carrier systems, the orthogonal multi-carrier transmission scheme offers increased robustness to mitigate wireless multi-path distortion effects, and any subsets of the available subcarriers can be flexibly assigned to any users according to their specific QoS requirements.This proposal seeks to explore effective solutions for managing radio and transmission power resources to guarantee the QoS performance as perceived by users with minimum transmission power consumption for emerging OFDMA-based broadband wireless access systems. We plan to design a cross-layer optimization scheme, where subcarrier and power resources are optimally allocated by jointly considering the information from both physical and upper layers. Information theory and advanced queuing theory will be combined together for modelling wireless system dynamics. Specifically, a novel channel estimation method is to be investigated for physical layer to estimate channel state information, and then to be utilized to formulate a robust power bit loading model. An advanced Markov Modulated Poisson Process (MMPP) queuing model is to be created for modelling QoS performance of upper layer heterogeneous multimedia applications. At the scheduler, a cross-layer multi-objective optimization will then be formulated and low-complexity algorithms are sought to search optimal solutions of joint subcarrier and power allocation. The outcomes of this project will make a significant contribution toward acceleration of the rapid and ubiquitous deployment of emerging next-generation broadband wireless access systems in the UK and worldwide.

Many organic molecules are delivered to us in crystalline form, ranging from foodstuffs such as the cocoa butter in chocolate, to pigments, propellants, and pharmaceuticals. Organic molecules can adopt a range of crystalline forms, or polymorphs, that have distinct properties, including melting temperature, colour, detonation sensitivity, and dissolution rate. This proposal will develop new ways of predicting and producing an extended range of polymorphic forms for a given molecule. Even when the molecule is not delivered in a crystalline form, a detailed understanding of its crystallisation behaviour is necessary for optimising the manufacturing process, and designing the product to prevent crystals forming (e.g. ruining a liquid crystal display). A major risk in the manufacture of organic products is the unanticipated appearance of an alternative polymorph, as resulted in the withdrawal and reformulation of the HIV medicine ritonavir, and of transdermal patches of a Parkinson's disease treatment that became unreliable once rotigotine re-crystallised unexpectedly on storage. Crystallisation is a two-stage process comprising nucleation (formation of stable clusters of molecules) and growth (growth of clusters until visible crystals are observed). The appearance of many polymorphs late in product development has been attributed to difficulties in nucleating the first crystals. However, changes in the impurity molecules present and contact with different surfaces may catalyse this nucleation. In this proposal we will explore the influence different chemical and physical surfaces have on nucleation of new polymorphs. Although many thousands of crystallisation experiments can be performed in developing a new product, this is costly and time consuming and it is impractical to test all possible conditions. Thus the ability to select specific predicted forms and design experiments to enable these forms to nucleate for the first time turns polymorphism into an advantage in product and process design. It would allow crystal forms to be selected and manufactured with the particular properties best suited to the intended application of the molecule. The research will also provide a deeper understanding of the true range of solid-state diversity that an organic molecule can display. The EPSRC Basic Technology program has funded "Control and Prediction of the Organic Solid State" which has established an internationally unique capability of predicting the range of thermodynamically feasible polymorphs for a given molecule. This project has demonstrated the capability to produce the first crystals of a distinctive new polymorph of a heavily studied anti-epileptic drug, by crystallising it from the vapour onto a computationally inspired choice of a suitable template crystal of a related molecule. This finding proves that totally new forms can be discovered using templates designed to target a particular computationally predicted polymorph. However, it is essential to understand the interplay between structure, surface, kinetics and thermodynamics in directing this process if we are to harness the underpinning science for wider applications. This interdisciplinary project seeks to establish the fundamental relationship between the predicted polymorph and the heterogeneous surface which promotes its formation. We will develop a range of methods for prediction and selection of likely polymorphs as well as novel crystallisation experiments and technologies, including inkjet printing. The detailed molecular level characterisation of how one crystal structure grows off another will produce a fundamental understanding of this phenomenon, allowing a refinement of the criteria for choosing the template. This will result in new experimental techniques and computer design methods that can be used to ensure that new organic products can be manufactured in in the optimal way without the risk of unexpected polymorphs appearing.

Abstracts are not currently available in GtR for all funded research. This is normally because the abstract was not required at the time of proposal submission, but may be because it included sensitive information such as personal details.