B2B software editor Cartesiam partnered with electronic services company éolane to develop Bob, a predictive maintenance assistant with embedded AI features which make it able to detect any equipment failures in real time, based on the machine vibrations.
According to both companies, “Bob acts as the “Shazam” of the vibrations”. It is made of vibration and temperature sensors which thanks to embedded machine learning algorithms allow it to “learn” autonomously and recognize normal equipment vibrations from irregular ones. The assistant then signals if it suspects any anomalies, whether it is a current problem or a possible future one.
When Bob spots a problem, it sends an alert to an IoT platform or a smartphone, but no production data to the cloud. For both partners, Bob is not meant to convey rough data to the cloud like a regular sensor would. Bob favors information which, thanks to its AI and calculation capacities, bring added value to the user.
EDF and Veolia Germany are among the first companies to have tested the solution. Bob will be commercialized in the next weeks, as a monthly subscription plan.