A Novel Sensor Based Approach to Predictive Maintenance of Machines by Leveraging Heterogeneous Computing
Tinku Malayil Jose · 2018
Predictive Maintenance anticipates equipment outages rather than reacting to them. This goes beyond preventive maintenance, which at least strives to shift machine downtime to noncritical periods. While the modern, advanced machines have sophisticated systems monitoring solutions integrated, most of the discrete or conventional machinery will not be having such facility and failure of these can cause disruption. To ensure that these non-smart machines are also monitored, the wise investment would be to fit devices which would help monitor them or else these are to be replaced with new/high end smart machinery. This situation demands a cost effective yet independent off-the-shelf sensor based device with easy integration and usage, which is introduced here, leveraging the capabilities of heterogeneous compute architecture, which can prevent unexpected machine failures by providing “right information in the right time” while eliminating “unplanned stops” thus bringing in absolute predictability on machine conditions.