Navigating the Connectivity Challenge in an IIoT World
What is the Connectivity Challenge for IIoT, and why will it persist?
The challenge is pretty straightforward … and fundamental. Simply stated, computing and connectivity obey different laws.
The law of computing is one of exponential power, Moore’s Law. From the dawn of computing to the present day, it’s the driving force behind Gartner’s recent projection of 20.8 Billion connected things by 2020. It’s also the reason 847 ZB of data will be created by 2021 according to Cisco’s Global Cloud Index. A relentless, exponential expansion of computing power (and price compression) in turn driving an exponential expansion of data everywhere.
The IIoT world simply would not exist without this law.
But what about the connectivity needed to make all of this computing useful? Sadly, the laws of networking are not the same. Connectivity requires infrastructure. And infrastructure obeys a different set of incentives involving things like backhoes and building permits and other real world factors such as cost/benefit analyses for any network build out. The same reason we never have enough roads during rush hour is the same reason we’ll never have enough networks to keep up with the onslaught of computing and the data that follows. In the IIoT world, the challenge is significantly worse as the incentives for high speed network investments at the edge or in remote parts of the world where IIoT is most often deployed are nonexistent.
In the face of this truth, how can the industry navigate the challenge and bring about the promise of IIoT? Here, as when faced with any scarce resource, the first step is efficiency.
The dominant architecture for most systems on the Internet is cloud computing, which involves moving all of the data into the centralized cloud. It’s straightforward and simple, but terribly inefficient, especially for IIoT data at the Edge.
Instead, successful IIoT deployments embrace reliable and efficient distributed communications protocols and more localized computing. Distributed protocols, unlike the point-to-point variants typically used when communicating with the cloud, are incredibly resilient and fault tolerant, ensuring effective and efficient communication even over unreliable connectivity at the edge. Furthermore, when IIoT data is completely isolated from any network, such as sensor data collected from disconnected oil wells, these protocols even automate previously manual “store and forward” techniques, allowing operators to effectively ferry large data out of isolated, “offline” IIoT networks into computing instances where the data can be utilized towards digital transformation.
The next step is an increasing reliance on Edge Computing that minimizes the distances these data need to travel. Taking connectivity out of the equation as much as possible means that IIoT analysis can occur much closer to where the data is being generated. As computing becomes cheap and ubiquitous, Edge Computing becomes increasingly practical as a computing model, but with many new management and operational challenges yet to be overcome.
Even so, effective Edge Computing must rely on the same distributed communications, given typical data management challenges are further compounded by the need to keep many more computing instances in complete sync across a variety of edge workflows.
The Key to Digital Transformation
Successful data management solutions for IIoT deployments must recognize the Connectivity Challenge and operate effectively within those confines. While IIoT is -the- key technology behind digital transformation, the promise of this technology rests squarely on successfully managing vast data across the Connectivity Challenge. Successful IIoT data management, in turn rests on more resilient protocols, designed for the unforgiving environment of the IIoT Edge.
No matter what, the Connectivity Challenge is here to stay.
At least until the laws of computing and connectivity converge. Perhaps when someone invents a backhoe that digs twice as fast, every 18 months, in accordance with Moore’s Law.
This article was written by Eric Klinker, the Co-founder and CEO of Resilio Inc, a distributed data management solutions provider for IIoT and Edge Computing. Prior to Resilio he was the President and CEO of BitTorrent Inc,. developing technology connecting more than a billion edge devices in the world’s largest distributed computing network.