In this article we meet Monolith AI, a deep technology company developing software tools that help engineers to design complex products like cars and aeroplanes. Their technology fuses computer science and physics in an artificial intelligence platform that can automatically validate and verify engineering designs without having to build physical prototypes.
Tell us about Monolith AI and what they do?
Monolith AI is a spin-off from the Imperial College London Uncertainty Quantification Lab, founded by Prof Francesco Montomoli and Dr Richard Ahlfeld (right and left respectively). We are developing virtual validation software for engineers using hybrid approaches between data analytics, machine learning and physical engineering simulations (CAE). We are working with large engineering companies in the aerospace, automotive and FMCG industries to deploy the technology through the product development cycle.
Monolith AI is funded by Imperial Innovations, Founders Factory, the Royal Academy of Engineering, the Digital Engineering and Test Centre and the Advanced Propulsion Centre. We were a finalist and named as one of 7 best Deep Science Start-ups in Industry 4.0 by Hello Tomorrow.
What challenges is Monolith AI addressing?
We are introducing machine learning and data science techniques to engineering to reduce the time and cost of bringing new products to market. The technology also reduces errors which improve product quality and further reduce the lifecycle cost of supporting products in service.
In industrial pilots, our technology has proven to solve complex CAE problems with up to one thousand times fewer simulations than traditional methods. This presents a significant opportunity to reduce the effort, time and cost when developing products.
Who are the organisations that Monolith AI work with?
We are already collaborating with an automotive manufacturer in the UK and aerospace and FMCG manufacturers in France. These companies are among the biggest names in their respective industries. We’re also working with several additional companies across aerospace and automotive about future collaborations.
What is Monolith AI hoping to achieve in the future?
We are developing modules that solve engineering challenges at each step of the product development lifecycle. Our vision is to consolidate all relevant engineering data into a single platform that supports engineers developing complex products. The platform will democratise knowledge such that all engineers will have access to the ‘experience’ of all other engineers in their organisation.
What accomplishments is Monolith AI most proud of?
Our technology was used to design a space launcher system that will soon be implemented. We identified the optimal structural solution with one thousand times fewer simulations than was previously possible, and we also identified ‘Black Swan’ moments that were previously missed.