The subject of the project is the industrial research and experimental development of a prototype device for multisensor data acquisition for preventive maintenance of machines, which will evaluate the behaviour of the machine and predict its need for maintenance or possible failure based on measured data, especially vibration, sound. The cloud-based software will further work with the measured data by using machine learning algorithm to evaluate the data. Over time, the algorithm will collect enough data for possible prediction of machine faults. Placing the algorithm on the cloud will enable high computational capacity and the ability to evaluate the data comprehensively. It is the sharing of data from individual machines, which may be of similar or the same type, that will enable the provision of more accurate and higher quality outputs over time in the form of maintenance planning or replacement of specific machine parts. Among other things, the possibility of powering these devices with a battery for easy installation will be explored.
Zdroje financování: Ministerstvo průmyslu a obchodu ČR - OP TAK – Aplikace.
