Don’t Push Me ’Cause I’m Close to the Edge

At a time when we all just started getting used to using terms like Industry 4.0 and the Internet of Things (IoT) in our everyday life, we get another batch of computer theory terminology to chew on.

Next up for the manufacturing world is edge analytics.

For those of us without advanced computer science degrees, the difference between edge analytics more and traditional analytics is location. Unlike the traditional process in which data is analyzed after being transmitted to a central location (such as a server or the cloud), edge analytics analyzes data at the collection point (near a machine’s sensors).

Computing power at the source analyzes data locally, eliminating the need to transfer data to a central computer or to the cloud. It produces the benefits of real-time data collection without the costs of bandwidth usage or data exposure risk.

Quite simply, the data collected and analyzed directly at the source gives manufacturers context about what is happening in the now, and does so quickly. Finding out if a part is out of spec can now be done even faster because the analysis is done near the point of manufacturing. However, for longer-term, operation-wide analysis, computing in the cloud likely remains the best choice.

In a new white paper counting down the top seven manufacturing predictions for 2017, Mark Watson, senior research manager for manufacturing technology at IHS Markit, identified edge analytics as a trend to watch.

“Throughout 2016, many cloud platforms were announced or released to support the Internet of Things (IoT) in manufacturing. While the remote cloud can offer significant advantages in terms of scalability and cost, concerns around cybersecurity caused hesitancy among end users,” he wrote.

What is likely to happen, though, is that data at manufacturing companies will become more secure, with the result being a combination of remote cloud computing and local edge computing being utilized.

To get to this best-of-both-worlds scenario, however, manufacturers must become even more computer-savvy.

Not sold yet? Well, the true value of edge analytics in a shop can be seen in two easily identifiable cases:

1. Data collected locally from machine tools quickly identifies errors, reducing the number of bad parts being produced because response time can be shortened.

2. Simulation and modelling can predict future machine failures. According to a recent Cisco/SCM World survey, manufacturing companies expect a significant 48 per cent reduction in unplanned downtime by using edge-based analytics systems.