Cambridge University to build AI chemical research lab
The University of Cambridge is to build a multimillion-pound research facility which will see AI accelerate chemical engineering progress.
While work has yet to begin on the build, projects have already started at the Innovation Centre in Digital Molecular Technologies (iDMT), part of the university’s department of chemistry.
It is hoped the facility will speed up access to pharmaceuticals, agrochemicals, functional molecules and molecular materias through machine learning and robotics-based synthesis.
The project was part funded by the European Regional Development Fund, and in partnership with pharmaceutical companies AstroZeneca and Shionogi.
Professor Alexei Lapkin from the University of Cambridge department of chemical engineering and biotechnology is iDMT’s director. He said, “Access to new functional molecules and materials continues to be a major bottleneck in many chemistry-using industries, such as medicine, food, electronics and energy.
“It is very difficult to predict how chemical processes would behave at an industrial scale. For this reason, development and optimisation of chemical processes usually takes quite a long time. AI tools can help solve complex problems of chemical process design speeding up the transition from a working chemical reaction in the lab, to a scaled-up industrial process.
“Combining cross-disciplinary expertise from several departments at the University, with state-of-the-art facilities and support from two of the leading companies in this area has the potential to enable the development of many new solutions for the nascent industry of digital molecular technology.
“Facilitating knowledge exchange to SMEs so that they can develop the right product offer that would serve the needs of the large end-user companies in the pharma, agritech and wider chemical manufacturing sectors will enable an industry-wide shift in how synthesis, process design and manufacture are carried out.”
'AI will free up scientists' time'
iDMT co-director and director of the EPSRC SynTech centre for doctoral training Matthew Guant added, “Despite tremendous advances in chemistry, we still cannot always make all of the molecules we need on demand, especially when set against increasingly competitive business-driven timelines, and this means that we often miss out on many potential opportunities to, for example, develop new medicines.
“The transformational change that we believe is required in the way chemical synthesis is approached is based on a radical increase in the throughput of chemical discovery and process development. This can be achieved through the automation of largely routine procedures, and the adoption of artificial intelligence to guide synthetic chemists towards successful solutions in a more efficient manner. This frees up time of a scientist to develop new ideas.”
Dr Ryuichi Kiyama, senior executive officer, Pharmaceutical Research Division, Shionogi, said: “We are proud to be part of this new innovative chemistry research consortium with the leading research institutes in the United Kingdom. As a pharmaceutical company with strengths in chemistry-driven small molecule drug discovery, we are committed to contribute to the discovery chemistry innovation in collaboration with researchers from the partner institutes and companies.”
The iDMT will support collaborative research projects with small and medium enterprises (SMEs) from across the UK, aiming to develop a technology base to support the emerging digital economy in the third largest manufacturing sector in the UK.
Three ways the iDMT will support research using AI
• Acceleration of synthesis through AI and automation
• Equipment for robotic experiments
• Algorithms and tools for digital process development
Google is using AI to design faster and improved processors
Engineers at Google are now using artificial intelligence (AI) to design faster and more efficient processors, and then using its chip designs to develop the next generation of specialised computers that run the same type of AI algorithms.
Google designs its own computer chips rather than buying commercial products, this allows the company to optimise the chips to run its own software, but the process is time-consuming and expensive, usually taking two to three years to develop.
Floorplanning, a stage of chip design, involves taking the finalised circuit diagram of a new chip and arranging the components into an efficient layout for manufacturing. Although the functional design of the chip is complete at this point, the layout can have a huge impact on speed and power consumption.
Previously floorplanning has been a highly manual and time-consuming task, says Anna Goldie at Google. Teams would split larger chips into blocks and work on parts in parallel, fiddling around to find small refinements, she says.
Fast chip design
They have created a convolutional neural network system that performs the macro block placement by itself within hours to achieve an optimal layout; the standard cells are automatically placed in the gaps by other software. This ML system should be able to produce an ideal floorplan far faster than humans at the controls. The neural network gradually improves its placement skills as it gains experience, according to the AI scientists.
In their paper, the Googlers said their neural network is "capable of generalising across chips — meaning that it can learn from experience to become both better and faster at placing new chips — allowing chip designers to be assisted by artificial agents with more experience than any human could ever gain."
Generating a floorplan can take less than a second using a pre-trained neural net, and with up to a few hours of fine-tuning the network, the software can match or beat a human at floorplan design, according to the paper, depending on which metric you use.
"Our method was used to design the next generation of Google’s artificial-intelligence accelerators, and has the potential to save thousands of hours of human effort for each new generation," the Googlers wrote. "Finally, we believe that more powerful AI-designed hardware will fuel advances in AI, creating a symbiotic relationship between the two fields.