Today’s enterprises are getting smarter, faster, and more cost-effective by letting IoT data guide their decision making. In factory environments, the availability of new, integrated sensor data enables better monitoring, situational awareness, and predictive maintenance—providing manufacturers with greater yields, improved safety, and reduced costs. However, for organizations to get the full benefits of smart, connected devices and operations, they must successfully merge their operational technology (OT) and information technology (IT) systems in order to rapidly ingest and analyze IoT data.
One approach is edge computing, where the digital world of IT systems and data processing exists on-site with the machines, sensors, controllers, and gateways within a factory. There are three reasons IoT strategies should start at the edge.
Connected devices in the manufacturing setting produce a different kind of data than what most IT organizations are familiar with – particularly high-frequency, time-sensitive data like temperature and pressure levels with a very short half-life. If it isn’t analyzed and acted on in near real-time, it quickly loses value.
OT systems with analytics and machine learning applications can use high-frequency data to better understand a process or machine, but failing to sample it at a high enough frequency can mean potentially missing anomalies with significant implications. By storing data locally, manufacturers can push more advanced analytics to the edge to predict things like machine failures before they happen.
As manufacturers add more connected devices to a network, the exponential increase in data will consume additional IT bandwidth. Moreover, high-frequency data typically translates into larger data sizes, pushing transfer and storage into the megabit and terabit range. This can intensify the strain on existing IT infrastructures—resulting in latency, degradation in service, or even outright loss of data.
As the cost of the processors and memory needed to implement IoT at the edge continues to decline, enterprise IT organizations can alleviate the issues of having to transmit large amounts of unfiltered data to a remote data center. IoT at the edge enables time-critical monitoring and analysis capabilities with more reliability, so manufacturers can send cleansed data to the cloud for better insights from their smart, connected operations.
Machine maintenance is generally based on historical data and requires careful downtime planning against production timelines. However, most data generated by factories currently remains unused, and data that does generate actionable insights is often based on spreadsheet analytics. As a result, the data collected becomes fragmented and siloed, existing in various incompatible formats that are inaccessible to IT.
Integrating data from disparate sources and formats in order to extract business value is one of the biggest challenges presented by the industrial IoT. Locating servers on the edge and connecting them directly to data sources and IIoT solutions such as ThingWorx enables new business insights which can reduce asset downtime, improve inventory management, and increase productivity.
If you’re interested in learning how enterprises can build and optimize an industrial IoT infrastructure using data from the edge, download this white paper.