Syngenta is bolstering its digital R&D infrastructure, signing a deal with TetraScience to connect its data sources and provide a backbone to its digital crop protection R&D capabilities.
The crop protection company will centralize its analytic data — including mass spectrometry and chromatography insight — with Tetra Scientific Data Foundry, which will structure data into an AI-ready format, according to a press release. Then, researchers can easily search for information through a central platform.
TetraScience scientists and engineers will implement the technology, working over time to help Syngenta’s research and IT team on the adoption of the technology.
“Delivering end-to-end data automation across our R&D organization requires a unified foundation - one that eliminates data silos, connects laboratory assets and systems, and transforms raw scientific data into accessible, actionable insight to drive the future of our science,” said Claudio Battilocchio, digital automation lead R&D at Syngenta, in a press release.
He added, “The capabilities provided by TetraScience offer that foundation, enabling us to standardize and harmonize data at scale across our R&D landscape. Such capabilities are fundamental to how we are transforming R&D —accelerating the speed and quality of scientific discovery, addressing productivity for data management, and ultimately strengthening our ability to develop the innovations that help farmers feed a growing world.”
At World Agri-Tech San Francisco, Syngenta revealed a partnership with quantum computing company QuantumBasel to explore the use of the technology in R&D, alongside existing AI capabilities. Feroz Sheikh, CIO and CDO at Syngenta, spoke with AgTechNavigator about the company’s digital strategy and using different tech tools to solve different problems.
“If we take any problem, whether it’s some of the research problems we are taking — like predicting chemical structures or image recognition problems — you can break down the problem into different parts. Some parts are still suitable [to solve with] traditional AI and neural network architectures. And some parts are more about predicting many different outcomes and then finding out what works best,” Sheikh shared during the video interview.
