Artificial Intellegens Newsletter (Sep 2023)

A new white paper summarises the results of an industry collaborative project funded by Innovate UK, coordinated by Intellegens, and with partners including Johnson Matthey, Domino Printing Sciences, Goodfellow, Welding Alloys Group, the Henry Royce Institute, and the Universities of Cambridge and Leicester. The project aimed to understand the current state of machine learning in the chemicals and materials sectors, assess the potential for and benefits of wider application, and identify factors that constrain adoption.

We had a great attendance at our recent webinar on machine learning for Design of Experiments (DOE). If you missed out, you can now catch the recording online. We compare ML with conventional statistics-based DOE methods and discuss how it offers the opportunity to go a step further, particularly for multi-component problems, systems with non-linear effects, or to enable adaptive DOE in which an experimental programme is refined as it progresses.

The Intellegens development team constantly looks for ways to make the Alchemite™ machine learning platform even faster. The combination of mathematical tricks in the underlying algorithms and practical improvements in the web interface make the system more responsive and user friendly. For example, the new ‘complete columns’ feature introduces speedups for datasets where the descriptor (‘input’) data is complete for every row, although outputs may be incomplete. This is common in experimental datasets where inputs are known but not every output is consistently measured. Handling denser, larger datasets more efficiently improves model training time, DOE and also analytic performance on the web platform.

In this month’s blog, we discuss the hype around AI for the life sciences, and chart a future in which machine learning becomes a routine tool – wherever there is data, there should be machine learning. Claiming “this drug was designed with AI” should seem as odd as it would be today to claim “this drug was partly designed with Excel”.

The Alchemite™ User Group Meeting will be held online on 17 October, and two of the user case studies that will be shared have now been confirmed. Clorox (chemicals) and FLSmidth (elastomers) will discuss what they are doing with Alchemite™. There will also be opportunities for the user community to share experience, learn tips and tricks, hear about the latest developments, and provide input to the Intellegens development team.

Thomas Bridge from our Science Team will present on “Harnessing AI for R&D in Drug Discovery” at the ELRIG-organised Drug Discovery 2023 event, Oct 18-19 in Liverpool.

We’ll be discussing machine learning for life science applications at TechBio UK in London (18 Oct).

We’re looking forward to discussing the strategic implications of machine learning for the chemicals industry at the CIEX 2023 event in Frankfurt (25-26 Oct).