T. Jesper Jacobsson

Bayesian optimisation (BO) is a sample‐efficient strategy for optimising expensive, noisy, black‐box functions, challenges that are common in experimental sciences where each experiment may involve significant cost, time, and effort. Despite its theoretical strengths, BO remains under‐utilised in experimental disciplines, in part due to its perceived mathematical complexity and lack of practical guidance tailored to non‐specialists. This tutorial provides a comprehensive yet accessible introduction to BO, with an emphasis on implementation in chemistry, materials science, and related fields. We cover the core components of BO, including surrogate modelling with Gaussian Processes, acquisition functions, and sequential decision‐making. In addition, we discuss practical extensions for handling noise, constraints, multi‐objective, and high‐dimensional optimisation, mixed‐variable spaces, batch processing, and computational scalability. Strategies for incorporating domain knowledge and deploying BO in automated experimental workflows are also presented. Throughout, we provide practical advice, highlight useful software, and emphasise conceptual understanding over mathematical formalism. As laboratory automation and data‐driven discovery become increasingly central to modern science, we argue that BO will play a transformative role in accelerating intelligent experimentation. The goal of this paper is to lower the barrier to entry and equip the reader with the knowledge needed to apply BO effectively in their own work.

Scientific discovery is severely bottlenecked by the inability of manual curation to keep pace with exponential publication rates. This creates a widening knowledge gap. This is especially stark in photovoltaics, where the leading database for perovskite solar cells has been stagnant since 2021 despite massive ongoing research output. Here, we resolve this challenge by establishing an autonomous, self-updating living database (PERLA). Our pipeline integrates large language models with physics-aware validation to extract complex device data from the continuous literature stream, achieving human-level precision (>90%) and eliminating annotator variance. By employing this system on the previously inaccessible post-2021 literature, we uncover critical evolutionary trends hidden by data lag: the field has decisively shifted toward inverted architectures employing self-assembled monolayers and formamidinium-rich compositions, driving a clear trajectory of sustained voltage loss reduction. PERLA transforms static publications into dynamic knowledge resources that enable data-driven discovery to operate at the speed of publication.

Accurately modeling the complex relationships among synthesis parameters, material compositions, and performance metrics is essential for accelerating the development of perovskite solar cells (PSCs). In this context, machine learning (ML) has proven to be a valuable tool. While most ML applications in PSC research rely on discriminative "black-box" models, this study adopts a generative approach by modeling the joint probability density function. We employ Gaussian Mixture Models (GMMs), a pragmatic and interpretable choice well-suited for the scarce, low-dimensional tabular data typical of PSC research. This single GMM framework is evaluated on five distinct tasks: discovering clusters, regression, generating novel configurations, training on data sets with missing data and, inverse design of the experimental (synthesis) conditions. That is, assuming we have the perovskite material composition and a target PCE, we infer the experimental conditions. For this latter task we use a novel "GMM-Assisted Optimization" method, which demonstrates to be more effective than standard random-start optimization, achieving an RMSE of 1.52 against target PCEs, more than halving the 3.32 RMSE of the baseline. These findings highlight the power of probabilistic modeling for data-driven discovery in PSC research.

This 6th annual Emerging PV Report surveys peer‐reviewed advances since August 2024 across perovskite, organic, kesterite, matildite, antimony seleno‐sulfide, selenium, and tandem solar cell architectures. Updated graphs, tables, and analyses compile the best‐performing devices from the emerging‐pv.org database, benchmarking power conversion efficiency ( PCE ), flexible photovoltaic fatigue factor ( F ), light‐utilization efficiency ( LUE ), and stability‐test energy yield (STEY) against detailed‐balance efficiency limits as functions of photovoltaic bandgap, and average visible transmittance ( AVT ) for (semi‐)transparent devices. Beyond efficiency, operational stability is assessed via degradation rates ( DR ) and t ₉₅ lifetimes. Highlights include single‐junction perovskite cells with efficiencies above 27%, organics surpassing 20%, and new Si/perovskite tandems exceeding 34%. Although multiple record efficiencies have been achieved this year, advances in mechanical robustness and operational stability remain inconsistent, especially in complex tandem stacks, emphasizing the urgent need for standardized protocols, improved large‐area homogeneity, and database‐driven benchmarks to accelerate the transition from laboratory demonstrations to scalable, real‐world deployment.

A plan for going from unstructured knowledge to structured and FAIR data with the vision of simplifying insight extraction and automated reproducibility.

Hybrid perovskites are interesting optoelectronic materials. The perovskite ABX₃ structure offers a vast compositional space, and we have identified over 300 perovskite ions. This flexibility enables tuneable properties and has significantly contributed to the success of perovskite optoelectronics. However, this diversity also leads to confusion, ambiguity, and inconsistencies causing challenges for data mining and machine learning applications. To address this issue, we propose guidelines and a JSON schema to standardize the reporting of perovskite compositions. The schema adheres to IUPAC recommendations and is designed to make data both human- and machine-readable. It captures key descriptors such as perovskite composition, molecular formula, SMILES representation, IUPAC name, and CAS number for each ion. To facilitate adoption, we have developed utilities to automatically generate comprehensive and standardized perovskite descriptions from standard ion abbreviations and stoichiometric coefficients. Additionally, we provide a curated database of all identified perovskite ions with associated descriptive data.

This 5th annual “Emerging PV Report” highlights the latest advancements in the performance of emerging photovoltaic (e‐PV) devices across various e‐PV research areas, as documented in peer‐reviewed articles published since August 2023. Updated graphs, tables, and analyses are provided, showcasing several key performance parameters, including the power conversion efficiency, open‐circuit voltage, short‐circuit current, fill factor, light utilization efficiency, and stability test energy yield. These parameters are presented as functions of the photovoltaic bandgap energy and average visible transmittance for each technology and application and are contextualized using benchmarks such as the detailed balance efficiency limit.

Material science has historically evolved in tandem with advancements in technologies for characterization, synthesis, and computation. Another type of technology to add to this mix is machine learning (ML) and artificial intelligence (AI). Now increasingly sophisticated AI‐models are seen that can solve progressively harder problems across a variety of fields. From a material science perspective, it is indisputable that machine learning and artificial intelligence offer a potent toolkit with the potential to substantially accelerate research efforts in areas such as the development and discovery of new functional materials. Less clear is how to best harness this development, what new skill sets will be required, and how it may affect established research practices. In this paper, those question are explored with respect to increasingly more sophisticated ML/AI‐approaches. To structure the discussion, a conceptual framework of an AI‐ladder is introduced. This AI‐ladder ranges from basic data‐fitting techniques to more advanced functionalities such as semi‐autonomous experimentation, experimental design, knowledge generation, hypothesis formulation, and the orchestration of specialized AI modules as stepping‐stones toward general artificial intelligence. This ladder metaphor provides a hierarchical framework for contemplating the opportunities, challenges, and evolving skill sets required to stay competitive in the age of artificial intelligence.