Compilers are central to computing, translating programs written by people into code for machines. Some aspects of compiler development, such as syntax analysis, bridge the theory and the implementation in a principled way, with lexers and parsers being algorithmically derived from high-level specifications. On the other hand, there is currently an unbridged gap between the theoretical specification of a programming language, given by a formal semantics, and the code produced by the compiler. Relating the two post hoc is possible, but difficult and rarely done. However, it doesn't have to be this way. A more principled approach is to begin with a semantics for the language, and seek to derive an implementation that is correct-by-construction. The investigators (Graham Hutton and Dan Ghica) have independently developed two such methodologies, which are based on complementary approaches to semantics (evaluators and abstract machines), but utilise different approaches to syntax (trees and graphs). The aim of this project is to reconcile the two methodologies to develop scalable and reusable frameworks for constructing certified compilers from semantics. The project combines the strengths of two leading research groups, is enhanced by a team of expert collaborators (Patrick Bahr, Mario Lavarez-Picallo, Edwin Brady, Simon Marlow, Anil Madhavapeddy, and Beniamino Accattoli), and is supported by fully-funded PhD studentships from the host departments.

Electech, covering areas such as sensors, power electronics, embedded computing, wireless communication technology, autonomous systems and large-area electronics, is predicted to play a foundational role in the future development of industries and value chains. It is central to Innovate UK's core strategy and its importance to future economic growth cannot be overstated. It is vital that the UK maintains a strong electronics design and technology base in the face of international developments. The proposed European chips act (February 2022), will mobilise 43 43 billion euros by 2030 in 'policy-driven investment' for the EU's semiconductor sector. The US CHIPS Act will result in a $280 billion investment to bolster their semiconductor capacity, catalyse R&D, create regional high-tech hubs and grow a more inclusive STEM workforce. The UK has a very vibrant but dispersed, electronic systems academic community, organised into larger activities in the universities of Glasgow, Imperial College London, Liverpool, Manchester, Newcastle, Sheffield, Southampton, University College London and Queen's University Belfast as well as satellite activities in a range of other universities. The community have been able to organise into an effective electronic systems community via the eFutures network (EPSRC eFutures2.0: Addressing Future Challenges grant, May2019-2023). In addition to growing the community, the objectives of the existing eFutures2.0 network had been to explore multidisciplinary opportunities for the sector. The successes of eFutures include: the organisation of 20+ in-person and online events (1825 attendees); the creation of a new website and a YouTube channel with 34 videoed talks (speakers from 19 countries) with a total of 1180 views; increased network membership by over 400% and move from a pure mailout model to include social media, achieving 64% of event attendees who had not previously engaged with the network; the delivery of two new, strategic landscaping reports: 'UK Landscape in AI & Brain-Inspired Computing Hardware' (Q4 2021) and 'Electronics for Healthcare: R&D across the UK' (expected Q1 2023). The 2021 Report had national media coverage, follow-up events (150 attendees), an upcoming, high-value proposal and a mention in the EPSRC Delivery Plan. The Healthcare Report results from online and in-person events (264 attendees) leading to a Programme Grant proposal. The network funded six multidisciplinary, concept projects (£78k), benefitting 11 academics across ten UK and four international universities; and delivered focussed engagement with 59 early-career and 30 mid-career researchers via two in-person workshops and online training. Ultimately, the aim is to further enhance the impact of UK electronics systems academic research and put the community in a strong, competitive position for collaboration with both national and international researchers, and industry. As highlighted above this will be achieved by continuing to build and growing network membership, organising the Net-Zero multidisciplinary event to engage our community more broadly in the area with other academic areas and companies to tackle this key topic, represent a strong focus on the electronics systems academic community in the UK, supporting early career researchers and growing the community by encouraging interaction or the national and international level and increasing the funding. We will achieve this by building on the successes of the eFutures2.0 activity with the same leadership team and steering group. The success and commitment to this activity is indicated by the in-kind commitment of £64,000 from our steering group companies.

Machine Learning (ML) already has a dramatic impact on our daily lives. ML developments in large language models and deep generative models cement that further. The recent explosion in ML, however, is built on the back of improved computer systems able to train and generate ever more powerful models. Systems design fundamentally defines ML performance and capability. This is true for Internet-scale ML and artificial intelligence (AI). Yet, more recently, it is especially evident in distributed, efficient, device-oriented, secure, personalised, privacy-preserving ML. UK strength in this fast developing area is dependent on a skilled R\&D workforce. Systems research and ML research are symbiotic. Current innovation in systems research is driven by the ubiquitous need for efficient and reliable ML. ML research, conversely, is steered by deployment capability and the economic and environmental impact of the resulting systems. Furthermore, systems research increasingly relies on ML methods to automate design, and ML research develops such methods. Major gains are made when the development of ML and systems are co-developed and co-optimized. This is relevant across a broad spectrum of industries: in-car systems, medical devices, mobile phones, sensor networks, condition monitoring systems, high-performance compute and high-frequency trading. Yet PhD training that brings together systems and ML is rare; research training is often siloed in the individual sub-disciplines. Instead, we need researchers trained in both fields and experienced in working across them. Hence: The ML Systems CDT will train a new type of student -- the ML-systems researcher. The ML Systems researcher is critically capable in both fields, and has collaborative research experience across the systems-ML stack. An example concretises this. A company is developing and deploying wearable body monitors. Effective models must be learnt on collected data, but data must be privacy preserving and bandwidth minimized. This is then personalised to each individual, adaptable to circumstance while being battery efficient and not connection dependent. To manage such a project requires knowledge of effective data-efficient ML signal analysis methods, designed and optimized for low-power hardware, itself tailored for the purpose through ML optimization methods. Knowledge of personalisation methods and the payoffs of privacy preserving methods vitally complement this. The societal impact, e.g.\ on those who might be obsessive about their medical state must also be considered, and will impact development. This CDT will train individuals with cross-cutting capability in all these components. Students must have broad understanding of different hardware designs, different platforms, different environments, different models, and different goals beyond their immediate research focus. This makes a cohort-based CDT vital. Standard PhD training in ML systems can result in research focus on a single ML technique and a single system. The CDT treats ML Systems as a holistic discipline. Cohort interaction, and integration gives students real experience across multiple systems, approaches and methodologies. Furthermore students will join together to contribute to a unified toolkit for the ML-Systems stack, and make use of others' contributions to that toolkit. On leaving the CDT, our graduates will understand fully where to focus resources to best improve a company's real-world ML development - whether that be at the ML-algorithm level, the hardware level, the compiler, level or even the legal level. They will be able to evaluate work at every level. We expect our graduates to be the leading team managers in real-world cutting-edge company ML.

Generative Models are AI models that can generate data. Recently researchers have shown that by training these models on large amounts of data (text data from the internet and images) these models learn to understand the regularities of our text and image world so well that they can generate responses to questions and create new images with surprising fidelity. This heralds a new era in which computers can assist humans to carry out tasks more efficiently than ever with significant opportunities for society, science and industry. However, these advances need significant research still -- how to make them train efficiently on different problems, how to understand their reliability and adherence to ethical norms.

Vision: to drive and promote advances in optical biosensing capable of translation to low-cost monitoring, and to build a broad UK community in low-cost sensing for healthcare. Precision medicine tailors healthcare to individual patient characteristics. We are now entering a new era of precision health, which shifts towards healthy individuals, asking how we prevent disease with appropriate interventions, prolonging healthy lifespans. New challenges include the urgent need for precise technologies to monitor individuals throughout life, and for improved methods to interpret this wealth of data. Precision health demands new physical biosensors that are low-cost but elicit rich biochemical information and can be used outside the clinic. This frees up clinician-time and focusses scarce resources. It is vital to develop methods to extract/exploit downstream patient-specific information from the sensors. Current exemplars ('BioSensors 1.0') are wearable devices (such as Fitbit, Apple watch), which record only superficial parameters (eg. temperature, acceleration, blood oxygenation), while glucose/insulin sensors provide only very specific data; the major challenge of providing comprehensive analytical information with an affordable portable device remains key for healthcare. The SARS CoV-2 lateral flow tests popularised the notion of personalised disease testing and showed it can be a reality however they lack sensitivity, reliable and consistent interpretation, and robust reporting capabilities. The leading groups assembled here have a track record of pioneering optical approaches for new paradigms in the biosensing domain, from conception through to market. Together, they propose to synergistically explore the underpinning fundamental science of 'BioSensors 2.0' and develop key demonstrators that address clinical needs while building a broader UK community of academics, SMEs, institutes, & clinicians to drive this paradigm to real demonstrators. Current portable sensors are too simple and limited in their capability. Instead, we need to translate advanced lab-based technologies into portable devices. Systems aspects need care, while miniaturisation is challenging. Sensors should achieve multiplexing, use machine learning algorithms to interpret outcomes, auto-calibrate to ensure long term operation, survive changing conditions, and attain small-enough limits of detection required for various biofluids. This is a time-critical juncture, as other countries will start to develop in this space, though nothing explicitly exists yet- the NHS as the main UK provider may be a great driver. We also focus on community building, with targeted activities to ensure the UK is placed to capitalise on sensor developments. Through building a Big Idea 'Making Senses' for the Research Councils across the wider Sensors ecosystem, our team identified with EPSRC the lack of UK leadership and joined-up academia-industry-govt networks. Engaging with a wide range of stakeholders from SMEs to large entities (NPL, CPI, LGC, Turing..) and multinationals (P&G, AstraZeneca,..), we find strong appetite and market pull for new types of biosensors with application domains beyond the hospital, as well as industrial settings. New ways to leverage light-matter interactions (in which the is UK internationally strong) for realistic biodiagnostics demands a broad interdisciplinary research focus. This confluence aims to develop entirely new industries of the future, and to energise the UK interdisciplinary science base, which is vital over the next 50 years as we realise the new paradigm of BioSensors 2.0.