Attendees at this month’s third Alchemite™ Focus Group reflected the growing and diverse usage of the machine learning technology, with representatives of sectors including chemicals, plastics, alloys, aerospace, paints, consumer products, additive manufacturing, energy, food and beverage, and printing technology.
Particular thanks go to BAT and Domino Printing Sciences for sharing their experience of applying Alchemite™ in product research and development. These presentations generated some useful ideas for other attendees, and discussion of factors including how to ensure effective adoption of machine learning where it requires changes to existing practices. In doing this, it was important to quickly generate and share results of direct benefit to the teams involved – for example, to provide chemistry teams with guidance enabling them to achieve their goals with fewer experiments, releasing time to focus on the more creative aspects of chemistry.
The Intellegens team presented recent and upcoming features. These included the ability to suggest initial experiments in scenarios where little or no data is available at the start of a project, and a variety of enhancements to analytics plots, enabling improved understanding of the machine learning model. A demonstration of a formulation design use case showed some of these features within a typical workflow.
Intellegens CSO Dr Gareth Conduit presented some interesting new science on extracting valuable information from what appears to be noise in experimental data. The research, focused on designing new concretes, was also discussed in this week’s Intellegens webinar – a recording is available here.
The meeting concluded with a discussion of future development plans in areas including image analysis and an invitation for all to continue the feedback and input process by working direct with the Intellegens team and attending future Focus Group sessions.
To keep in touch with everything going on at Intellegens and be informed of future events, subscribe to our newsletter.