Getting Ripe, Delicious Blueberries to Your Plate

From creamy yoghurt topped with ripe big blueberries, to fluffy blueberry pancakes and muffins, blueberries are a healthy and popular fruit worldwide. But getting those blueberries to you does take quite a bit of planning, logistics, and smart math and analytics.


Hortifrut has been a worldwide leader in berries since the early 80’s. They grow blueberries in several locations in Chile, Peru, Mexico, and other countries, and distribute them to partners and clients worldwide. You can imagine this is not an easy task, as sending a fresh fruit worldwide in a cost-effective way requires managing production fluctuations that depend largely on unpredictable events such as weather and pests, while accounting for shipping costs, market demand and price, and customers’ quality requirements. Gonzalo Bustos, head of data analytics at Hortifrut puts this in perspective, “Our planning process requires understanding a myriad of scenarios. For example, a cold spring in Chile could delay production, reducing access to key markets that are willing to pay higher prices, but this gap could be hedged if we could fulfill that demand from another country with different weather patterns, such as Peru. Solving a problem of this complexity required a deep understanding of Hortifrut’s operations, data and processes. We selected EpiX Analytics to help us, based on their long-term track-record in using advanced analytics to create value and achieve strategic goals.”


EpiX worked closely with Hortifrut to capture this information, prioritize features, and then implement a comprehensive optimization model called BlueSquid, that optimally matches the production of blueberries with demand worldwide to maximize Hortifrut’s sales or gross profits. To do this, the analysis considers many variables and constraints, chiefly:

– The types and quality grading of blueberries that have different demand in different markets. For example, some markets buy primarily organic berries whereas others focus mostly on the quality grading.

– The duration and costs of shipping by air or sea.

– Weekly (i.e. seasonal) production of berries on each site, and demands on each market, paired with the uncertainty in the prices at destination which can fluctuate wildly during the year.

The optimization model was developed using Frontline’s powerful Analytic Solver, allowing Hortifrut to have easy and full access to all functionality, logic and results in a user-friendly tool that allows them to easily update all model inputs to perform scenarios on the fly. The optimization was defined as a linear problem, giving Hortifrut a guarantee that the globally optimum solution will be found within a few minutes.


Hortifrut has been using BlueSquid constantly to plan for its upcoming growth season under multiple scenarios. Moreover, the model is being actively used to examine the effect of various strategic initiatives under scenarios such as changes in regional demand, opening of markets currently restricted for quality or sanitary reasons, or increasing production in other parts of the world to offset production fluctuation risks.

“We are not just seeing huge savings in the time that it takes to plan production, distribution and sales, but we’re actually now able to plan for higher sales, and better understand the value of various potential strategic initiatives,” says Gonzalo Bustos. “BlueSquid is delivering value from the planning department to the board level.”


Training - Food Safety for FSA

The UK Food Standards Agency (FSA) needed effective and efficient remote training for a diverse…

Modeling health impacts of diets: a risk analyst's perspective

In this 2-part webinar series we present nutritionist-friendly explanations of the modeling…

Webinars on COVID-19

Our COVID-19 webinars review key lessons in probabilistic modeling: Looks at different models that..