“Amongst one of our biggest challenges was to build the right infrastructure to enable faster, more efficient research that contributed to the sustainability of our planet. It needed to be stable and resilient to scale and yet be technically efficient to match the fast development cycles of RStudio. The solution needed to fully understand the requirements of our scientists so they could effectively collaborate and help farmers adapt to climate change, soil erosion and biodiversity loss and other business critical applications”
Luis Galiano, Senior Lead, Statistical Analysis and Data Visualisation at Syngenta.
As the world’s largest agricultural innovator, the Syngenta Group uses data science to transform how crops are grown by enabling millions of farmers to make effective use of available resources. Through world class science and innovative crop solutions, their 49,000 people in over 100 countries play a vital role in safely feeding the world, whilst taking care of our planet.
Central to Syngenta’s strategy has been their Good Growth Plan, which places the fight against climate change and biodiversity loss at the heart of farming’s productive future. Their innovative use of statistical analysis is to transform crop protection and maximise genetic gain, whilst seeking ways to produce food more sustainably worldwide.
With digital agriculture shaping Syngenta systems across the globe, increasing emphasis was placed on enabling and automating the extraction of knowledge from data to optimise important decisions and support this step-change towards regenerative agriculture. The team wanted to centralise an R ecosystem for their community of 75 users with an emphasis on improving the data quality, processing, and ongoing support to a wide range of functions. Crop science is a justifiably highly regulated industry, and a centralised ecosystem offered an opportunity to bring standardisation and the dissolution of siloed data applications and code repositories.
Enabling the delivery of packages more efficiently and easing the extraction of data was a fundamental challenge. They required a solution that allowed stakeholders to innovate whilst having centralised access to data from multiple resources and the ability to develop and deploy customised analysis and visualisation solutions, faster.
As users with a heavy dependence on R, Syngenta were keen to have a flexible consultancy engagement, that would enable a common platform for open application development in R. “It was essential to create the right, resilient platform suitable for the community of users. We wanted to allow our scientists to scale a proof of concept in an open way,” said Luis.
It was this need for a scalable R ecosystem that led Syngenta to engage Mango as consultancy partner on the design of an appropriate modern technology platform, since their introduction of R as a proof of concept into the business in 2013. After in depth discussions with users the desired platform would underpin Syngenta’s data-driven transformation, preventing data silos and inefficient practices. A centralised RStudio ecosystem presented the right solution for Syngenta to enable the most appropriate collaborative development amongst their users. It also enabled the management of their vast datasets at scale, in a controlled reproducible way.
The desired infrastructure provided accessibility to the large and diverse team, providing them with the desired end to end capability i.e., databases, API’s, to apply custom code and algorithms, create product ID’s, interfaces and develop applications and publish internally for stakeholders, all with internal ownership.
“The value to our Scientists has been considerable. We wanted the opportunity to find what we did not know. To build an infrastructure and an open sand box, enabling innovative research that aligned to our sustainability goals,” said Luis. Helping to drive Syngenta’s sustainability agenda, the centralised platform has leveraged multiple business applications of R and Shiny, automating regulatory documentation and process, enhancing data quality, and saving considerable time and resource, enough to free 2 PHD scientists. The team have optimised sample logistics, business case modelling, field conditions and crop production modelling which otherwise would not have been possible without a centralised platform. They have successfully developed coherent frameworks and best practice, compliant methodology in R. “It has improved the productivity of scientists, allowing us to focus on the needs of each grower in the markets we serve,” said Luis.
“We now have an agile proof of concept to grow organically with our business goals. Our business centric approach has been a justification of a quick win for Syngenta and added incremental business value leading to better experimental design,” said Elisabeth Easton, Crop Protection R&D IT Informatics Platform Lead.