D’isum announced the publication of a paper that introduces a failure prediction by capturing subtle precursor phenomena that appear a few days to a few weeks before the failure, with monitoring the vibration and sound of machinery equipment, preventing failures in advance. The key features of this approach are as follows:

It is a highly simple and cost-effective approach that captures precursor phenomena solely from normal data without requiring training data. Since normal data can typically be acquired within two to three days, the system can be operationalized quickly. It can be referred to as a “today’s failure prediction” method. This solution does not require advanced software skills or data analysis expertise. This method is effective not only for machinery equipment but also for early detection of anomalies in activities such as “driving a car” or “abnormal health conditions”. You can download the document from here (link to a PDF page).