We’re proud to announce the launch of Crystal Ball! This exciting new feature allows you to start crop quantities based on your forecasted harvest.
When starting batches in Artemis, you plan how many plants (or units) you intend to start for each batch. You can always reconcile the amount when actually planning. Now, using Crystal Ball, you’ll also see your forecasted outputs, taking into account the forecasted loss percentage for each batch.
Our data science team predicts generated outputs using a variety of methods, and only the best, most accurate forecast is shown. That means you’re always getting the most accurate picture of what’s expected to happen throughout production. Don’t worry, if you don’t have enough data to predict the outputs accurately, you won’t see Crystal Ball for that output.
We also include a range for our prediction. This is especially helpful if you have contractual obligations for a given quantity of units. Let’s say an order is placed for 10,000 lbs of product for example; our forecast may show your current planting to generate 10,000 lbs, but the range may be between 8,000 and 12,000 lbs. In that case, you may choose to increase your seeding quantity to bring the minimum up to the required contract amount.
Another great thing about this feature is that Artemis will predict all generated outputs for the batch. For example, if you generate off the same batch, both raw weight and product for clamshells, Artemis will tell you how much of each you can expect.
Predicted loss is also shown alongside the forecast. This is particularly important if you’re taking into account loss from stage to stage, seasonality, and other factors causing crop loss in your cycles. We show you how much loss we expect, as well as a range to account for variability. The loss is taken into account in the forecast.
At this time, Crystal Ball is only active when a template has a single harvest point. It does not support crops with cut-and-come again harvests or scheduled partial harvests.
Coming soon: we’ll be adding tooltips to tell you which of our prediction models has been used to forecast your outputs.
If you want to learn more about our platform and how you can get more accurate forecasts, reach out to firstname.lastname@example.org and one of our reps will get in touch!