The fascination with vibrations and the phenomena measurable through them gave rise to the development of a piezoelectric sensor and, quite literally, paved its way into existence. This innovative sensor found application in the far north within the Aurora Intelligent Road Project initiated by the Transport Agency. Utilizing machine learning, this sensor distinguished between cars based on their measurements, evolving to a point where it achieved a remarkable level of precision, even recognizing pedestrians with an accuracy approaching that of well-known automobile brands.
A noteworthy incident during the sensor's operation occurred when uncertainty arose about one of the pedestrians that was detected. Upon closer inspection, a video from the measurement site revealed that the identified pedestrian was, in fact, a squirrel crossing the road. This amusing revelation underscored the remarkable sensitivity of the sensor and hinted at the diverse possibilities it offered.

In these contemporary times, we have come to recognize the vast potential of machine learning in predictive maintenance. Previously absent from our toolkit was the abiliti to predict equipment failures and forecast their timing. Today, our system is equipped to alert us to the smallest deviations detected in a device's operation. It not only monitors the unfolding situation but also provides an opportunity for timely repairs. This proactive approach allows for the scheduling of maintenance downtime during periods of low criticality, mitigating the impact of possible production interruptions.

In the world of Innocode, your device now has statuses beyond just running and stopped.