Smart Factories and the Challenges of the Proximity Network
comes off the cut line and recording this information for real-time monitoring by the plant manager( Figure 3).
Figure 3: A Gateway-Mediated Edge Connectivity and Management Architecture at a Chicken Plant
2.2 Predictive Maintenance
While the majority of factories explore how Machine-to-Machine( M2M) and IIoT technologies improve process, few factories have been able to successfully implement predictive maintenance. Many factories have mechanisms for scheduled preventative maintenance and redundancy on critical systems, but few predict failure in time to perform maintenance. Predictive maintenance generally involves collecting sensor data such as temperature and vibration from critical components( such as motors) and analyzing it over time. The algorithms for detecting maintenance conditions can be as simple as threshold crossing or as complex as a trained neural network.
We have noticed two prominent predictive maintenance business models emerge. In one case, the plant integrates sensors from their critical equipment into their own systems and manages all of the data and networking themselves. In other cases, the factory consumes a piece of equipment from a vendor such as a motor in a compressor. The vendor seeks to expand its business model by offering a service to monitor its equipment.
IIC Journal of Innovation- 65-