Cropster, a software solution for the coffee sector, has published the newest version of its Roasting Intelligence, which features bean temperature and Rate of Rise (RoR) predictions. It predicts the next two minutes of bean temperature and the RoR by using Artificial Intelligence.
With Cropster’s new technology, roasters can see where their roast is going next before it gets there. The company has created a machine learning algorithm that takes gas changes, and the machine type into account to predict the bean temperature curve.
Machine learning is a base form of ‘Narrow Artificial Intelligence’, meaning that it uses algorithms based on large data sets that is employed against a specific use case. Cropster is using the large quantities of available data it can process and has developed proprietary algorithms to make inferences and predict outcomes in near real time.
The company said that its bean curve predictions would help roasters develop new profiles and train new staff members more quickly.
According to Martin Wiesinger, CTO of Cropster Cropster worked on technologies that pushed the specialty coffee industry forward. The company concentrated its efforts on building a system around data that is readily available to its entire customer base to bring the benefits of this transformational technology to everyone who uses Roasting Intelligence.
Christian Ott from Mission Coffee Co said that Cropster’s early beta testers reportedly had positive feedback about the new bean curve predictions.
According to Cropster, bean curve prediction begins 60 seconds into the roast and appears as a dashed curve extending past the running bean temperature curve. It updates every second, which means the further into the roast people go, the more accurate the prediction becomes.