Research Fellow – School of Mechanical Engineering – 106924 – Grade 7 Full-time

Research Fellow Position

The FAST (Formation and Ageing for Sustainable Battery Technologies) project is a major Faraday Institution consortium led by the University of Birmingham with partners across Oxford, Cambridge, Warwick, Nottingham, Imperial and UKBIC. Its mission is to transform the battery formation and ageing stages—currently the most time-, energy- and cost-intensive steps in lithium-ion cell manufacturing—by building a scientifically informed and scalable framework for next-generation production.

A key workstream of the FAST project provides the digital and analytical backbone of the programme. It develops sensor-enabled diagnostic cells, multi-modal data pipelines and hybrid physics-informed machine learning approaches to understand interfacial behaviour during formation and to optimise process protocols. The Research Fellow will play a central role in this work.

The post holder will design and implement data extraction and preparation pipelines for heterogeneous datasets spanning electrochemical testing, embedded sensors, environmental logging, spectroscopy and advanced imaging. They will create and curate structured, FAIR-compliant datasets suitable for multivariate analysis and machine learning, ensuring high-quality metadata, traceability and reproducibility.

Building on this data foundation, the Fellow will develop hybrid modelling tools that integrate physics-based insights with data-driven methods—such as physics-informed neural networks, surrogate models and Bayesian optimisation—to explain formation behaviour, identify early indicators of cell performance, and reduce reliance on empirical testing. Working closely with engineers, modellers and experimentalists, they will generate interpretable, scientifically grounded models that directly inform the design of improved formation and ageing protocols.

The Research Fellow will:

• Work within the FAST research programme, delivering the data engineering and modelling tasks that underpin Workstream 1b, and contribute to preparing project reports, presentations, and future funding proposals.
• Operate within the specialist area of data engineering, machine learning (ML), and physics-informed modelling, applying advanced computational methods to heterogeneous battery formation datasets generated across the consortium.
• Analyse, interpret, and integrate multi-modal research findings—including electrochemical time-series data, imaging outputs, embedded sensor measurements, and environmental logs—to create structured, interpretable, and reusable datasets that support hybrid modelling.
• Contribute to generating funding by co-authoring sections of new research proposals, demonstrating how data workflows, digital infrastructure, and ML approaches can support emerging research directions in battery manufacturing, diagnostics, and sustainable engineering.
• Contribute to pathways for commercial translation, including opportunities for software tools, modelling frameworks, or data pipelines to feed into licensing, future spin-out activities, or industrial adoption by partners such as Agratas, Volklec, the UK Battery Industrialisation Centre (UKBIC), Gaussion, Illumion, and Oxford Battery Developments.
• Support public understanding and dissemination of the discipline by contributing to open-source data standards, transparent modelling documentation, FAIR (Findable, Accessible, Interoperable, Reusable) datasets, and accessible explanations of physics-informed ML models for academic and industrial audiences.

Main duties include:

• Data Engineering & Preparation
• Develop automated pipelines for ingesting, cleaning, and structuring data from sensors, electrochemical testers, imaging systems, and environmental logs.
• Establish metadata standards and ensure datasets meet FAIR principles (findable, accessible, interoperable, reusable).
• Create high-quality, ML-ready datasets through feature extraction, multivariate analysis, and robust quality control workflows.
• Modelling & Machine Learning
• Develop hybrid models that combine domain physics with data-driven techniques (e.g., PINNs, surrogate models, Bayesian optimisation).
• Work with experimental partners to interpret formation signatures and validate model outputs against real-world measurements.
• Build interpretable, mechanistically grounded models to identify early predictors of formation success and inform protocol optimisation across the consortium.
• Research Communication & Collaboration
• Publish research outcomes in high-quality journals and present findings at scientific conferences, consortium meetings, and workshops.
• Engage closely with interdisciplinary academic teams to align data workflows, modelling approaches, and scientific objectives.
• Support and mentor PhD researchers and students working on aligned data or modelling tasks.
• Industry Engagement & Impact Delivery
• Work with industrial partners to ensure data pipelines and modelling tools reflect real manufacturing needs.
• Translate project outputs into formats useful for commercial stakeholders, such as interpretable dashboards, technical briefings, and data summaries.
• Participate in industry advisory board sessions, technical review meetings, and collaborative sprint activities that test emerging modelling insights in industrially relevant settings.
• Contribute to creating demonstrable impact pathways, supporting the transfer of modelling tools, datasets, and protocols to commercial partners.
• Project Coordination & Governance
• Help maintain the shared FAST data infrastructure, documentation, and reproducibility standards across all partner institutions.
• Contribute to discussions on data governance, ethical data handling, and best practices for multi-partner research.
• Actively contribute to the University's and project's commitment to equality, diversity, and inclusion, fostering a collaborative and supportive research environment.

Person specification:

• Essential Qualifications
• PhD (or near completion) in engineering, computer science, physics, applied mathematics, or a related discipline with a strong data-driven or machine learning focus.
• Essential Skills & Experience
• Strong programming skills (Python essential) and experience with scientific computing libraries.
• Experience building or working with complex datasets, ideally from experimental or sensor-based systems.
• Hands-on experience with machine learning, including deep learning, probabilistic models, or physics-informed approaches.
• Ability to analyse, visualise, and interpret complex time-series or multi-modal data.
• Strong communication skills, including the ability to explain technical concepts to non-experts.
• Ability to organise own research, manage priorities, and collaborate effectively within a diverse team.
• Desirable Experience
• Experience implementing ETL (Extract, Transform, Load) pipelines or working with data management platforms.
• Familiarity with Bayesian optimisation, surrogate modelling, or scientific ML.
• Background or interest in modelling physical systems (electrochemical or otherwise).
• Exposure to FAIR data principles or scientific database design.
• Experience working in interdisciplinary teams or coordinating with experimental researchers.