A lot of the buzz these days in AI is about genetic AI and how AI is being used to accelerate software and consumer products. Today, an AI startup called Physicsco-founded by two theoretical physicists – including a Formula 1 engineering superstar – emerges from stealth with a very specific focus on building and operating physical systems in the business world.
London-based PhysicsX has built an artificial intelligence platform to create and run simulations for engineers working in project areas such as automotive, aerospace and materials science manufacturing – industries where there are regular development bottlenecks due to the way which models are tested before production. It comes out of stealth today with $32 million in funding.
The round, a Series A, is led by General Catalyst. Others in the round include a very interesting mix of financial and strategic backers. They include Standard Industries, NGP Energy, Radius Capital and KKR co-founder and co-executive chairman Henry Kravis. The funding will be used to grow the business and continue to develop the company’s platform. This is PhysicsX’s first external funding.
PhysicX addresses a problem that has been very consistent but ignored in the world of manufacturing and physical production.
In any physical system, whether in an experimental laboratory or a live industrial environment, whenever a new idea is introduced—say, a theory to improve the operating efficiency of a machine, not to mention work on entirely new products—engineers you need to simulate how the new idea will work before you commit to its development and further refine how it works. Typically, this simulation and testing work is conducted by scientists, engineers who may use some artificial intelligence in the process, but ultimately edit the process by hand.
“Something like airflow in an object might take you an hour or two, but if you want to simulate something more complex, it might take you a day or more. So there is a computational cost and therefore a time cost to this. And that limits the depth to which you can optimize,” Robin Tuluie, who co-founded PhysicsX with Jacomo Corbo, said in an interview.
The couple knows the pain points firsthand all too well.
Tuluie already had two different lives as a theoretical physicist. As an academic, he worked alongside Nobel Prize winners with a focus on astrophysics. He then moved into the world of racing, first at Renault and then at Mercedes, respectively as Head of R&D and Chief Scientist, where he devised designs that helped his teams win four Formula 1 world championships (winning some reputation himself at edit, process). He also spent years at Bentley and Volkswagen working on car design.
Corbo, who received his PhD from Harvard, has also worked in racing, but most recently founded and ran QuantumBlack, the artificial intelligence labs at McKinsey, working with several Formula 1 clients as well as other automotive and industrial clients. in thorny product engineering problems.
The pair assembled a team of at least 50 scientists — mechanical engineers, physicists and others — to create the PhysicsX platform, which addresses the automotive industry but also a much broader range of applications, Corbo said.
“We’re building an enterprise platform to support a fairly broad range of domain applications that are connected to manufacturing and optimization problems, physics simulation bottlenecks,” he said. “What PhysicsX buys you is the ability to predict physics [of a system] with very, very high precision and fidelity, doing it anywhere from 10,000 to a million times faster. Now be much more sophisticated about, say, mining, in a very high-dimensional space.”
The appearance of PhysicsX comes at a very timely moment in the world of deep learning and artificial intelligence, especially in how it applies to the physical world.
Just earlier this month, DeepMind released new research on how it was applying highly advanced machine learning to the world of short- and long-term weather forecasting, and Corbo believes the natural shift will highlight the next frontier of AI research and development.
“This is the first time that artificial intelligence models, these deep learning models, these geometric deep learning models, have outperformed numerical simulation for weather,” Corbo pointed out. “We’re starting to see this happening across physics more broadly. And, and that enables a lot of different applications in the engineering space, so we’re building a platform to be able to do that across all domains and across a wide range of domain problems.”
Businesses, in general, have faced many obstacles when it comes to digital transformation — tearing down existing infrastructure to adopt more modern IT technologies and approaches. While you can classify what PhysicsX is doing as a kind of “digital transformation,” the startup can sidestep these challenges because the kind of applications it faces, in engineering and R&D, aren’t typically IT issues that require scaling organizations wider.
Still, it’s a new approach that will disrupt the way industrial companies approach development today. Therefore, General Catalyst is betting on a very hot area – artificial intelligence – but also breaking new ground by supporting a startup that believes how this hot area will evolve.
“PhysicsX is expanding the boundaries of engineering in critical areas, led by a team deeply skilled in simulation engineering and machine learning,” said Larry Bohn, MD of General Catalyst. “With reliability, customer relationships and technical expertise, we believe PhysicsX is poised to transform engineering in complex industries. This aligns with our vision for industrial transformation and gives PhysicsX the opportunity to build a class-defining company in advanced industries.”
