Inside McCain Foods’ Plan for 100% Regenerative Potatoes

Modern farmers are navigating unprecedented market volatility, driven by erratic climate patterns, escalating input costs and global geopolitical tensions. To safeguard the global food supply, agricultural systems must evolve from surviving to thriving. 

Regenerative agriculture offers a robust pathway toward long-term resilience and profitability, yet achieving widespread adoption requires scalable, data-backed solutions built for the dirt, not just the laboratory. 

As an enterprise deeply rooted in agricultural heritage, McCain Foods has committed to implementing regenerative practices across 100% of its global potato acreage by 2030. 

Standardising imperfect field data

In the world of precision agriculture, decision-support tools are frequently celebrated for their eco-efficiency. 

Software innovations are already proven to significantly reduce environmental footprints, with some systems demonstrating capability to reduce nitrogen fertiliser use by up to 53% without sacrificing crop yields. However, the ultimate test is scalability. 

Deploying sophisticated software across massive, geographically diverse farming operations quickly exposes severe integration bottlenecks between cloud-based models and legacy tractor hardware.

Dr. Michelle Lynn D’Souza, Leader in Research and Innovation of Global Agriculture at McCain, explains that the firm’s foundational approach bypasses proprietary isolation in favour of open synergy. 

“Our digital twin is designed to integrate with existing GIS-enabled equipment to simulate real farm conditions to model regenerative practices, test outcomes and guide decision making,” she says.

Yet, achieving this level of harmony is easier said than done. The reality of commercial agriculture involves managing massive data fragmentation.

According to Michelle, the true technical hurdle lies in the infrastructure itself: “The challenge is that farm data is often inconsistent, fragmented and unreliable. 

“Data varies in quality across hardware, often manually ingested due to legacy systems and lacking standardisation, with added risks around commercial sensitivity. 

“These complexities are compounded globally, where varying conditions, from climate and heat to power and connectivity, affect performance.”

To overcome these localised limitations, McCain’s engineering strategy relies on establishing flexible protocols rather than demanding pristine, unrealistic infrastructure. 

“To address this, we are establishing a ‘minimum viable data standard’ that defines thresholds for quality and frequency so tools can operate reliably with imperfect, real-world data across diverse global farming systems,” Michelle adds.

Building a replicable tech stack for commercial scale

A key pillar of this global strategy is McCain’s Farms of the Future initiative. 

Operating as live innovation testbeds across Canada, South Africa and the UK, these sites – alongside 30 global innovation farms – allow the company to trial controlled-traffic systems, new crop species and advanced data architectures. 

Backed by significant digital agriculture investments alongside academic institutions like the University of New Brunswick, the project acts as a blueprint for scaling regenerative farming.

From a tech-stack perspective, the primary challenge is synthesising inputs from wildly different sources – such as localised soil sensors, satellite imagery and biodiversity trackers – into a unified platform that a standard commercial grower can actually afford to replicate. 

McCain’s philosophy intentionally avoids forcing growers to buy into hyper-expensive, high-density hardware networks.

“We are building a modular, replicable data stack anchored in widely available inputs – satellite imagery, weather models, yield monitors and GPS-guided equipment – with more advanced sensors as optional layers,” Michelle emphasises. 

“Rather than requiring high-density soil sensor networks, which are not always feasible, we focus on interoperability and accessibility.”

This modular architecture allows the technology to meet farmers where they currently are, rather than where a software engineer wishes they were. 

By deploying this framework, McCain allows growers to calculate their own digital maturity curves. Michelle highlights how this approach translates directly to the field:

“This enables a practical ‘gap analysis’ allowing growers to benchmark their current data and hardware, identify missing components and calculate the investment required to close those gaps to scale incrementally. 

“The result is a realistic blueprint that reflects the constraints of commercial farming while still enabling robust, data-driven decision-making.”

Designing for failure in a living sandbox

In pure software engineering, edge cases are tested in isolated sandboxes. In agritech, the sandbox is a living ecosystem vulnerable to torrential rain, extreme heat and heavy machinery vibrations. 

For Michelle’s team, early prototypes and unexpected field failures have proven to be the most valuable teachers.

Since 2018, McCain has collaborated with Dalhousie University’s Faculty of Agriculture to support field-based innovations on their Canadian Farm of the Future. 

One notable pilot involved upgrading standard farm sprayers with RTK-GPS and high-resolution cameras to scan fields for insect pests at an individual plant level during routine field operations. 

While the concept looked perfect on paper, the physical environment of an active farm quickly pushed the hardware to its limits.

“One lesson from this prototype was mounting cameras on a sprayer boom,” Michelle reveals. “It worked in controlled conditions, but in the field, constant movement caused mechanical damage and data loss. Other issues, like overheating hardware and limited space for GPUs, reinforced how unpredictable farm environments are.”

This rough-and-tumble reality completely reshaped how McCain approaches its core digital twin hardware assumptions. Instead of building for a flawless stream of metrics, the systems are now engineered to survive chaos. As Michelle puts it: “The lesson is clear: design for failure. 

“Systems must assume data dropouts, physical stress, with modular hardware and validation layers that distinguish between true data gaps and hardware disruption. We are baking these learnings into the design of the digital twin hardware assumptions.”

This philosophy underscores a massive conceptual shift in how the tech industry views digital twins. 

In a manufacturing plant or an automotive factory, a digital twin models highly controlled variables with extreme precision. When mapping an open-air farm, however, you are dealing with highly complex, living biological systems where predictability is nonexistent.

“In a farm environment, unpredictability is the standard, not the exception,” Michelle notes.  “Unlike factories, where variables are controlled and models can be highly precise, regenerative ecosystems involve biological processes that introduce constant variability.”

Because of this inherent volatility, McCain’s digital twin doesn't chase the mirage of a flawless mathematical forecast. Instead, the focus shifts toward agility. 

“Our digital twin focuses on actionable insight rather than perfect prediction – combining precise modelling where possible with validated heuristics where it’s not,” Michelle says. “The value lies in giving farmers real-time visibility of emerging changes and enabling earlier intervention in a dynamic, living system.”

Transparency from soil to shelf 

The ultimate goal of accumulating this massive dataset isn’t just to help farmers optimise their inputs, it’s also to connect the entire supply chain. 

Today’s consumers increasingly demand ecological transparency, a trend McCain is capitalising on with initiatives like its ‘Taste Good. Feel Good.’ campaign, which connects sustainable farming directly to consumer choices. 

Behind the scenes, this requires transforming raw agronomic data into verifiable, auditable pipelines that can back up environmental claims.

“We’re integrating on-farm data into a global, standardised dataset with auditable dataset, linking real farming practices to consumer-facing transparency,” Michelle says.  

“This underpins campaigns that help consumers feel confident about their planet-friendly choices.”

Yet, asking farmers to change how they manage their soil involves significant financial risk. 

This is where financial technology (fintech) becomes a crucial piece of the puzzle. To de-risk the transition for McCain’s network of 4,400 growers, data architecture must evolve to satisfy financial underwriters and risk modellers, creating an ecosystem where sustainable practices are directly rewarded with capital.

Michelle believes fintech will be a powerful tool for spreading these practices around the world. 

“Fintech could play an important role in scaling adoption – quantifying ROI, modelling risk and enabling outcome-based financing,” she concludes. 

“By combining data, transparency and financial tools, we can reduce upfront investment barriers and support farmers in transitioning to more regenerative systems. 

“Ultimately, scaling regenerative requires these elements working together from soil to shelf.”