Accelerating materials discovery using artificial intelligence, high performance computing and robotics

Pyzer‐Knapp, Edward O.; Pitera, Jed W.; Staar, Peter; Takeda, Seiji; Laino, Teodoro; Sanders, Daniel P.; Sexton, James; Smith, J. R.; Curioni, A.

doi:10.1038/s41524-022-00765-z

Cited by 138 publications

(104 citation statements)

References 67 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[ 58 ] Text mining techniques can extract unstructured data hidden in published documents (e.g., papers, patents, datasheets, and reports) and turn them into structured materials data. [ 77 ] This extraction procedure can be done via different chemistry‐aware NLP toolkits such as OSCAR4, [ 78 ] tmChem, [ 79 ] and ChemDataExtractor. [ 80 ] Huang et al [ 81 ] via using ChemDataExtractor on 229 061 academic documents extracted a battery database that contained 292 313 data.…”

Section: Limitations Of Artificial Intelligencementioning

confidence: 99%

Artificial Intelligence‐Based Material Discovery for Clean Energy Future

Maleki

Asadnia

Razmjou

2022

Advanced Intelligent Systems

View full text Add to dashboard Cite

Artificial intelligence (AI)‐assisted materials design and discovery methods can come to the aid of global concerns for introducing new efficient materials in different applications. Also, a sustainable clean future requires a transition to a low‐carbon economy that is material‐intensive. AI‐assisted methods advent as inexpensive and accelerated methods in the design of new materials for clean energies. Herein, the emerging research area of AI‐assisted material discovery with a focus on developing clean energies is discussed. The applications, advantages, and challenges of using AI in material discovery are discussed and the future perspective of using AI in clean energy is studied. This perspective paves the way for a better understanding of the future of AI applications in clean energies.

show abstract

Section: Limitations Of Artificial Intelligencementioning

confidence: 99%

Artificial Intelligence‐Based Material Discovery for Clean Energy Future

Maleki

Asadnia

Razmjou

2022

Advanced Intelligent Systems

View full text Add to dashboard Cite

show abstract

“…Bringing NLP together with other powerful parsing and collection techniques, it is possible to build comprehensive datasets of the available knowledge for a specific domain like the synthesis of silicates [37]. Other recently published perspectives highlight the details of these developments from a technological point of view [38,39].…”

Section: Building Datasetsmentioning

confidence: 99%

Developing Catalysts via Structure‐Property Relations Discovered by Machine Learning: An Industrial Perspective

Joshi

Wilde

Asche

et al. 2022

Chemie Ingenieur Technik

View full text Add to dashboard Cite

Industrial catalyst development is a complex issue that requires optimization of performance, synthesis, costs, and engineering aspects. During the development, structure-property relations are often used to provide valuable insights into the catalyst. However, conventionally, this process is time-consuming and costly. Advancements in the field of automation for experimentation, data collection, and simulations have allowed the use of machine learning (ML) strategies for this development. Herein we provide an industrial perspective on ML strategies for the development of solid catalysts.

show abstract

“…When working on searching large discovery spaces for material candidates, we encountered high reuse of sub-graphs across multiple virtual-experiments. To handle this, we introduced a prototype approximate memoization scheme (Section 3) for computational workflows which helped us accelerate the search [34,37,47]. We found this was facilitated by the abstractions we had introduced which made it straightforward to implement suitable single-node equivalence criteria.…”

Section: Approximate Memoizationmentioning

confidence: 99%

“…We trialed this method on the set of six virtual experiments we developed in [34]. There were three distinct experiments each with two alternates (see Figure 2.1).…”

Section: Automatic Computational Workflow Compositionmentioning

confidence: 99%

Towards an Approximation-Aware Computational Workflow Framework for Accelerating Large-Scale Discovery Tasks

Johnston¹,

Vassiliadis²

2022

Preprint

View full text Add to dashboard Cite

The use of approximation is fundamental in computational science. Almost all computational methods adopt approximations in some form in order to obtain a favourable cost/accuracy trade-off and there are usually many approximations that could be used. As a result, when a researcher wishes to measure a property of a system with a computational technique, they are faced with an array of options. Current computational workflow frameworks focus on helping researchers automate a sequence of steps on a particular platform. The aim is often to obtain a computational measurement of a property. However these frameworks are unaware that there may be a large number of ways to do so. As such, they cannot support researchers in making these choices during development or at execution-time.We argue that computational workflow frameworks should be designed to be approximation-aware -that is, support the fact that a given workflow description represents a task that could be performed in different ways. This is key to unlocking the potential of computational workflows to accelerate discovery tasks, particularly those involving searches of large entity spaces. It will enable efficiently obtaining measurements of entity properties, given a set of constraints, by directly leveraging the space of choices available. In this paper we describe the basic functions that an approximationaware workflow framework should provide, how those functions can be realized in practice, and illustrate some of the powerful capabilities it would enable, including approximate memoization, surrogate model support, and automated workflow composition.

show abstract

Accelerating materials discovery using artificial intelligence, high performance computing and robotics

Cited by 138 publications

References 67 publications

Artificial Intelligence‐Based Material Discovery for Clean Energy Future

Artificial Intelligence‐Based Material Discovery for Clean Energy Future

Developing Catalysts via Structure‐Property Relations Discovered by Machine Learning: An Industrial Perspective

Towards an Approximation-Aware Computational Workflow Framework for Accelerating Large-Scale Discovery Tasks

Contact Info

Product

Resources

About