New Data Project Tackles Potato Storage Losses Across North America

Cellar Insights expands its predictive data program to help growers spot spoilage early and protect crop value.

An insulated commercial potato storage warehouse filled with a massive bulk pile of potatoes. A red conveyor system is active, and a single worker stands in the foreground monitoring the operation. This image visualizes the large-scale industrial storage environment that Cellar Insights is optimizing with its Rotten Potato Data Project to reduce spoilage using predictive analytics and machine learning.
Large-scale commercial potato storage facilities, like this one, are vulnerable to significant economic losses from spoilage. The Rotten Potato Data Project by Cellar Insights aims to use machine-learning models and predictive analytics to identify rot risks weeks before visual symptoms appear, helping growers stabilize their inventory and reduce overall storage losses across North America.
(Photo courtesy of Cellar Insights)

Cellar Insights, an ag-tech company that provides smart remote monitoring and predictive analytics systems for commercial potato storage, has launched its Rotten Potato Data Project to better assess storage losses and improve early-warning tools throughout North America.

The project, which launched with growers and storage operators in the Canadian provinces of Manitoba, Alberta, Prince Edward Island and New Brunswick, expands to the U.S. this month.

Spotting Spoilage Before It Spreads

Ross Culberson, a sixth-generation potato grower, co-founded the company to help improve the quality of perishable crops in storage. Ryan McDonald, vice president of strategy with Cellar Insights, says it was in those early days that they “realized quickly that everybody still had to go walk their bins and inspect their stuff partly because of the severe risk of spoilage.”

So, spoilage became a key focus for the company. McDonald says that Cellar Insights homed in on conditions, sensor readings and what those measurements could indicate in terms of spoilage. Through the company’s work, it has detected about 90% of spoilage issues, tracking potatoes from the moment they enter storage, McDonald says.

“What we want to do is give lead time,” he explains. “We look at the conditions that correlate to spoilage, so it’s a difference between rot risk and rot presence. We can start to give a rot risk profile saying that it’s not necessarily a significant amount of spoilage yet, but you’re headed in that direction.”

McDonald says even when high risk correlates to rot presence, the early indicators show rot presence a few weeks ahead of visual scouting by a grower.

“Typically, we were three or four weeks advanced [rot] presence indication ... before a grower could say with any confidence that there’s a problem,” he says, noting the system has strong tracking to see if rot progresses past that initial indication or not.

Turning Storage Data Into Actionable Risk Models

Showing growers that rot can stabilize with the right intervention provides immense value, McDonald says.

“Then they can hold through it for their contract and get the full value for the extended storage that they had planned with their processor or buyer,” he says.

Growers and storage operators who enroll in the project share digital records of past spoilage events — including rot, breakdown and hot spots — with the company to help develop a library of storage events and match the events to sensor and controller data that preceded them. Cellar Insights says the data provided will help train and evaluate machine-learning models designed to identify storage risk earlier and more accurately.

McDonald assures growers that the data involved in the project is owned by the growers and is confidential. Cellar Insights anonymizes all data before use, and specific operations or events are never shared or published. The company says it uses the information to strictly to train storage-risk models, with zero impact on commercial relationships.

“We’re really taking data privacy seriously and the management of who gets to see what by helping that supply chain,” he says.

Incentivizing Grower Participation Across North America

McDonald says having more growers participate helps build a better system.

“No crop’s the same every year. No storage is the same. Farmers run their storages differently,” he says.

Cellar Insights wants to move beyond just reliability of detection to offer growers more awareness in the type of spoilage and the potential impact.

McDonald says, for example, “If you’re reducing your humidity, what [will] the trade-off point between shrink and saving ... be? Those are all things that we don’t officially do now, but through more data and more learning, we hope to be able to get to that level.”

Cellar Insights compensates growers for quantifiable submissions. McDonald says the compensation helps Cellar Insights gather as much data as possible. Growers can also donate the compensation to charity. There’s also a big discount for participating growers to Cellar Insights’ system the next year.

“We wanted to be able to reward growers for that information,” he says.

Building a Connected Potato Supply Chain

McDonald says growers, too, will start to understand more of the root cause of spoilage events by looking at temperatures, humidity and all the other potential factors.

“There are lots of times we’d have spoilage events, and it was clear in our system, but then trying to go through all the details is something that … was a bit difficult to get all the details,” he says. “By putting a little more incentive and a program around it, we’re finding it’s much easier for everybody to give a more complete set of information that we can use to do a better job and continue to do a great job going forward.”

McDonald says one U.S. customer does about a billion pounds of potatoes a year, and Cellar Insights has been able to import all the data into one place to track different varieties or products.

“We bring all the information into a single view so you can compare across your storage and for a specific variety or specific product,” he says. “You can see how your inventory and supplies are doing and make some really solid improvements to your logistics in your supply chain.”

McDonald says there’s also growing interest in this type of data beyond just growers to have a more informed supply chain and make better decisions.

“We found that the growers are more and more willing to share that information so that everybody can benefit through the whole supply chain,” he says.

The Packer logo (567x120)
Related Stories
Early reports from the field point to strong quality across both onion and potato crops with harvest expected to continue through October, says the Idaho Falls, Idaho-based grower-shipper.
With planted acres down across major growing regions, suppliers look ahead to stabilized pricing alongside major operational upgrades and steady consumer demand.
Kwik Lok outlines how integrated closure label printing helps suppliers meet strict traceability demands, streamline packaging operations, and boost consumer engagement.
Read Next
New data show the cyclospora outbreak is driving a widespread pullback across the fresh produce aisle — even for non-recalled categories — triggering the sharpest single-month retail lettuce price drop in BLS history.
Get Daily News
GET MARKET ALERTS
Get News & Markets App