The current industry practice for yield prediction is a time consuming, labour-intensive and expensive process of sampling. Small samples are taken from the vineyards during the growing season and extrapolated to determine overall yield. The accuracy of the yield estimated varies significantly and have severe impact on the business.
Our yield prediction model uses the power of machine learning, 50+ data points from weather sensors, ground sensors and historical information of the plots to predict the quantity of grapes that will be harvested from a plot under observation at the end of the season.
It starts with your yield estimation model utilising your plot's historical data and then augments your model with real time data to create a unique model for your vineyard.
The resultant model is self learning and, learns from a variety of sources and data points locally and globally, therefore, is always iterating on its own to give you the most accurate information on your yield through out the season. It will add at least 15-20% accuracy strength to your current model used for yield prediction.
This information allows you to get a base data to build on and plan. It supplements your decision making on yield and makes the process faster, simpler and incrementally accurate.