Converging at the IoT Edge

By Alon Segal

As the edge becomes the heart of the Internet of Things, there is a need for true expertise to win through.

We're seeing a great deal of interest in Internet of Things (IoT) hardware from software players as the true importance of IoT edge computing emerges. Google has announced its proprietary edge processing unit, Microsoft has its own edge microcontroller units, and Intel is teaming up with Dell Technologies to deliver IoT solutions—three giants looking to capitalize on the shift of intelligence to the edge.

This focus on the edge is being driven by the incredible rise in IoT intelligence and control moving into the "things" themselves, which is breaking down the conventional cloud model. ABI Research forecasts that three million new IoT edge devices will be added each day by 2023, and the industry is poised to change dramatically now that NB-IoT networks are being rolled out. Furthermore, the evolution of mobile infrastructure is not just about adding bandwidth and capacity—it is also about correcting modes of support in order to accommodate billions of things with different profiles of data behavior.

Re-balancing the IoT means that connected things at the edge need to become smarter—in short, they need to do more of the work, be more autonomous and work congruently with the network. This means that instead of a deluge of unfiltered data being pumped to a remote cloud, information sent from the edge must be locally processed and analyzed. For example, a heating, ventilation and air conditioning (HVAC) controller need not constantly update an ambient temperature reading; instead, it should only communicate changes, thereby conserving the energy of any battery-operated devices and making best use of constrained bandwidth such as NB-IoT.

Networks are more efficient when raw data has already been transformed into results in near-real time. This increase in clustered intelligent activity closer to the edge has been termed fog computing in the past, but may require a new definition as it moves into the devices and things directly to the next-generation IoT edge.

This emerging edge is created by the combination of an increase in the computational power of edge devices; advances in distributed software technologies, tools and standards; and the ability to bring these all together in future connected products. At the same time, the resulting products need to meet the dropping price points, size and power-consumption requirements necessary to fulfil the "manufactured into everything" imperative for the IoT industry.

However, delivering effective IoT solutions is not just about manufacturing hardware. The industry is evolving rapidly, and a full understanding of real-life, environment-driven constraints and demands of connected things is key to being able to deliver intelligent IoT modules that can power this edge convergence. Nor is it simply possible to get there from the software side of the convergence—hence, the appetite by the large software giants to extend into the cellular-connected hardware world is not native for them, nor is it where they have true domain expertise.

The edge is becoming the heart of the IoT, where an enormous range of things live and work, from tiny sensors embedded in logs to combat timber theft to huge machines supporting distributed manufacturing processes. However, on the module side, many of the newcomers—and, in particular, those from countries with low manufacturing costs—are doing nothing more than executing reference designs from the cellular chipset vendors into module-level products. The DNA is not there for either the software players or the low-cost vendors to catalyze the conversion successfully.

The IoT will be the largest collection of connected devices the world has ever seen—and it will take expertise and experience to deliver it successfully and efficiently at the edge and in the cloud.

Alon Segal is responsible for Telit's technology roadmap and product strategy. He is a visionary technologist and entrepreneur with global work experience. Throughout his career, he has held senior technical roles that contributed to business growth and leading-edge innovation. Prior to joining Telit, Alon was the chief product officer and co-founder of Zoeticx, a groundbreaking SaaS middleware solution that turns health-care IT into an active environment, where patients' medical information reaches out for providers' attention. Alon also served as the VP of portfolio product management at Amdocs, the CEO at Telespree Communications and the general manager for production printing systems at Electronics For Imaging. He has authored and been granted multiple patents, and has been recognized as a recipient of numerous awards. Alon received an MBA degree in finance and operations research from Tel Aviv University and a B.Cs. cum laude degree from Technion (Israel Institute of Technology).