PepsiCo is one of the largest potato buyers in Europe. Of the 3,900 farmers supplying directly to the snacks and drink giant across Europe and sub-Saharan Africa (ESSA), approximately 1,200 are potato farmers.
These potatoes end up in a number snack brands, including Walkers and Lay’s crisps, whereas procured corn, oats and peanuts feed into PepsiCo’s Doritos tortilla chips, Quaker Oats, and Duyvis peanut products.
The snacks and drink giant claims a “heritage” of securing local crops for local production. “Spanish potatoes are set to the Spanish factories, and the Portuguese potatoes to the Portuguese factories,” PepsiCo Europe’s head of agricultural procurement, David Wilkinson, told FoodNavigator.
Blending procurement traditions with new technologies, PepsiCo has developed precision agriculture methods to ensure resources are efficiently used across all farming operations.
“The challenge we, and all industry, are facing at the moment is [centred around] climate volatility,” Wilkinson continued. “The precision agri and farm tech solution sector is going to become so important in helping us better manage these volatile situations.”
PepsiCo continues to build on precision agriculture methodologies. Indeed, its 'next-gen' iCrop tool is currently being applied to all 500,000 hectares of potato farms in the ESSA regions. “At the end of the day, [the tool] is all about providing insights for different crops to our growers, and help them grow better.
“Our aspiration is to roll out this technology across all our corn crops, our peanut crops, and our oat crops,” we were told.
Data points inform iCrop app
Precision agri tool iCrop was born out of a collaboration between PepsiCo and the University of Cambridge almost a decade ago. The idea was to track the performance of new potato varieties and develop variety-specific agronomy programmes with farmers. Today, the ‘next gen’ iCrop tool comes in a mobile smartphone app that links to a secure data server online.
The technology relies on capturing approximately 240 data points per field. These points record, in real-time, the soil type, field location, potato variety, the date of planting, and the date the crop emerges. “We are then able to disseminate that information to support our farmers with in-field decision making,” Wilkinson explained.
More specifically, data captured throughout the growing season will help PepsiCo better plan its longer-term sourcing strategy with its growers. “We determine which are the best varieties for them, what the best soil types are for those varieties, whether we should plant earlier or later, whether they irrigated in the best way, could we use fewer inputs, or could require more inputs,” he continued.
“All of these precision agri elements [help us to] provide much more information to the farmers.”
Improving irrigation methods in Spain
iCrop has helped PepsiCo's potato growers improve on farming practices across a variety of areas, including in water management. In Spain, for example, the new iCrop tool is supporting irrigation scheduling to ensure that water is used as efficiently as possible.
And the progress is clear. Having monitored its irrigation in 2012-13, PepsiCo determined that in just 25% of cases, the right amount of water was being applied at the right time. In 2015, the firm became the first to implement an irrigation scheduling tool for potatoes, and increased that irrigation efficiency to 48%, Wilkinson recounted.
“Last year, be applying the irrigation tool on a much more bespoke basis – on the basis of every single field, different variety performances, and using exact planting dates – we were able to increase the efficiency of our water being applied to 92%.”
This means that 92% of all fields had the right amount of water applied when the crop needed it most, he continued.
Looking ahead in the short-term, PepsiCo plans to continue building expertise in satellite intelligence and Nvidia data to “alternative how we capture crop information, provide more real-time data points to the growers in a more efficient way”.
Artificial intelligence and machine learning may well become part of the story, he continued. “As we start to develop more data points and more multi-year data across all the crops, I think that machine learning and analytics will be able to play a bigger and bigger part.”