Whether you want to create smart insurance contracts, know how long your belongings have traveled, by what means of transport or if they have suffered any shocks, you can now develop embedded machine learning models with breathtaking precision.
Before the arrival of ML, recognition and classification of movements were very hard and time-consuming tasks for embedded developers. It was necessary to go through complex decision-tree algorithms based on the activation of specific thresholds at different time intervals. Now it is a completely different story, as we can build custom solutions with only a small amount of data in a record time.
During The Things Conference Embedded, Louis Moreau showed how to do that using Edge Impulse and The Things Industries' Generic Node. Watch his session below!