We recently talked to Pete Bartolik of IoT World Today about the various edge computing frameworks that enterprises have to choose from. He outlines efforts from LF Edge, OpenStack and the Eclipse Foundation in this article.
Here are some additional thoughts in response to questions about some of the efforts underway:
Q:There are multiple open edge projects underway, what impact is that going to have on organizations trying to figure out where to invest in edge infrastructure and how to go about deploying edge devices?
Many companies, especially those engaged in IIoT initiatives, don’t view themselves as using edge computing, at least not in the sense that they are moving to open platforms built around modern microservices-based architectures. For those organizations, some will figure out a strategy with existing vendors showcasing their participation in a particular project. Others will find their way based on the work of employees, some of whom might not be full-time developers, but are just looking to solve a particular problem. Those folks will do some research and download the code they need, and provision applications on something as basic as an Arduino board. To that end, the standards efforts need to do the same things that a successful vendor does – market the solution at events, host developer hack-a-thons and basically work on publicizing what they do.
Q: The open projects that seem to be getting the most attention are EdgeX Foundry and other projects under LF Edge, Eclipse Kura and ioFog, and OpenStack StarlingX. Are these competing efforts or are they ultimately coming together in a rational manner?
These efforts have some overlap, but also areas where they are complementary. Eclipse’s new Edge Native Working Group is going to focus on creating awareness of its orchestration platform, for instance, which meshes with the microservices that have been built under the auspices of EdgeX Foundry. Enterprises still either need to bring these components together or rely on vendors and integrators to help them.
Q: I realize some of these efforts were started in vendor labs and are probably already impacting deployable products, but what’s a realistic timeframe for buyers to gain access to products that can truly be considered the standard, open products they can use in their edge architectures?
I think there are already elements of a complete edge compute stack that is open in the sense that they are built on open-source code. There is still a lot of custom work needed to bring all the elements together.
Q: What role are the hyperscalers (AWS, MS Azure, Google Cloud) playing in facilitating or complicating the reality of open edge frameworks?
They are doing both, I think. On the one hand, they want to be seen as facilitating open edge efforts and will talk about using open-source software in their edge service portfolios. That being said, their natural tendency will be to make it frictionless to stay within their ecosystem (or hard to leave, depending on your point of view), so buyers will need to evaluate things like code portability and whether a single vendor approach will be adequate for their needs.