Analytics “at” the edge, on the device, in real time

It takes time for information to travel from the edge of the Internet of Things ( IoT ) to the cloud and back. That’s what happens every time a node takes a measurement, sends it to the cloud to do some sort of analytical calculation, then sends it back to the edge for an action to occur. Unfortunately, there are instances where this is not a viable scenario, including those in the medical, automotive, and military markets. With Industrial IoT (IIoT) hitting the mainstream, the need for a real-time response is growing quickly.

If you could do those calculations at the edge and avoid going up to the cloud except when absolutely necessary or when there’s a down time, you could save in various areas, not the least of which is the cost of sending that data back and forth, possibly through a cellular carrier.
You may think such a scenario would be difficult to implement, but fortunately it’s quite the opposite thanks to a technology partnership between Wind River , a global leader in delivering software for the IoT, and Greenwave Systems , a managed services company. The two companies have done the hard part for you, as they integrate Wind River’s real-time operating system (RTOS) with Greenwave’s IoT analytics tools.

The partnership results in a very tight integration between the RTOS and the analytics. It’s designed such that any product being developed with VxWorks can be made more intelligent, with most decisions being made as close to the product as possible, minimizing the need to go out to the cloud.

As people understand the cloud, there’s a common debate about how to segment the workload between it and the edge. Obviously, it depends on your application, configuration, environment, use case, and so on. We’ve found that there’s a real need to have real-time control close to the edge so there are no latency or data loss ramifications.
One example of an industrial application that can benefit from such technology is a wind farm comprised of a series of wind turbines. Typically located in remote places, including miles out at sea or in a desert, the cloud is usually accessed over a cellular link or an even more costly satellite connection. That high cost usually negates the use of a real-time link.