Home » Technology » Vinsight gives grape and almond growers a high-tech crystal …

Farmers know how hard it is to predict yield accurately. High-value crops like grapes or almonds prove particularly challenging. Farmers expect, and settle for, twenty to forty percent uncertainty with these crops. Vinsight founder and CEO, Megan Nunes, who grew up in the farming business, found that unacceptable.

After almost a decade-long career in the satellite industry, she turned her tech knowledge back home. Her Redwood City, Calif.-based startup now offers forecasting software and data analytics to farmers who are growing almost anything that’s not corn, wheat or soy. While Vinsight did not have permission to name its early customers, the CEO says, it is already working with one of the world’s largest wine making businesses and the world’s second largest producer of almonds.

Rather than installing its own proprietary sensors or weather stations in the filed, or scanning farmland from on high, Vinsight amasses all the data that it can from growers, government agencies and other sources. It analyzes the data in aggregate, and comes to understand when optimum yield is likely or not, based on a wide variety of correlations.

For example, even if you experience a perfect balance of sunny days and rainy ones, if windspeed is high enough, grapes in your vineyard will shatter and won’t be salable. Or, if your decidedly more durable almond trees bloom too quickly, they won’t yield the maximum amount of nuts they could. In an ideal weather year, that could be due to bee flight hours and a lack of pollination.

Nunes said Vinsight has hit a 10% error rate, which is a step in a positive direction, and three-times better than industry standard. Here’s how they make it happen, she said. “What we do is constantly analyze and take in data on a daily basis, including from remote sensors, weather stations and satellites, to identify what is happening on a farm as it correlates with crop performance or yield. We also look at data trends on a ten to…

Source link