FOLLOW

FOLLOW

SHARE

Does The IoT Need Edge Computing?

In order to manage the huge influx of data it will bring, new infrastructure is needed

7Jul

The Internet of Things (IoT) is already changing the world in a myriad of ways, and we are yet to even really get started. Gartner estimates that there will be some 8.4 billion connected devices installed worldwide by the end of 2017, up 31% on 2016, with roughly 37% of these devices set to be used by businesses and the rest by consumers. By 2020, they say there will be more than 20 billion connected devices.

The explosion of data from these devices presents a tremendous opportunity to companies. Those ready to analyze it will be rewarded with insights into every facet of their day to day operations. However, there a number of challenges that need to be overcome first. One of the most important is the ability to process this data, with cloud computing under pressure to meet the data computing and intelligent service demands of IoT devices and applications.

There is, however, a solution increasingly being touted - edge computing. In cloud computing, processing power is centralized. Data must travel from a device to servers to be processed, with the output then pushed back to the device. Edge computing, on the other hand, pushes generation, collection, and analysis out to the point of origin - the IoT devices and sensors - as opposed to a data center or cloud.

Ian Scales compares edge computing to the human nervous system, writing on Telcomtv that: ‘There’s an important component in human physiology called the Autonomic System which more or less does for us humans what edge computing is designed to do for the cloud. The relevant part of the Autonomic ‘system’ is the ‘sympathetic nervous system’. This is the thing a doctor is tapping into when he or she hits you on the knee with that little hammer - it’s designed to trigger your ‘quick response mobilizing system’. This allows you to jump into action without actually engaging the brain to think about it first.’

Edge computing is perfect for IoT for three reasons. Firstly, because the data is processed nearer the point of origin, you reduce the latency between devices and data processing layer, thereby enabling faster response and decision making. This also means that the costs related to ingesting a large amount of data in the cloud are considerably lower and network capacity is freed up for other workloads.

Secondly, as edge computing means the data is localized, should any individual device malfunction, it does not have a knock-on effect on others as a result. Retaining the data locally also provides a boost to compliance and security as there are fewer opportunities for hackers to access all data at once.

Finally, by distributing and storing your data into smaller data repositories, you can more easily segment manufacturing analytics into specific types of manufacturing and geographic regions as you don’t have to pull data extracts from a centralized corporate database. This makes data aggregation considerably easier and allows you to provide real-time analytics directly to managers in specific regions.

This potential is now being realized by a number of the major tech companies, including Microsoft’s Azure IoT Edge service and Huawei’s Edge-Computing-IoT (EC-IoT) Solution. Amazon has also now gotten involved, with the recent announcement at its re:Invent conference in Las Vegas that it was making AWS Greengrass generally available. As noted on VentureBeat, ‘Greengrass lets customers write functions that can be deployed on compatible devices and run in response to triggers from local hardware or the AWS cloud. Using those functions, it’s possible to handle data processing without a network connection, which is key for IoT devices.’ It can run on-premises, has a small footprint that can run on system-on-a-chip devices like Raspberry Pi and BeagleBone powered by ARM processors, as well as in full-blown x86 servers in large data center environments. The compute layer of Greengrass is delivered through AWS Lambda, the first server-less computing platform.

Dima Tokar, the co-founder and CTO of MachNation, noted that: ‘For prospective customers evaluating cloud platforms, the general availability of Greengrass will give AWS temporary advantage over Azure, as it provides a production-ready set of capabilities for edge devices. For hardware vendors evaluating cloud platforms, Amazon shipping Greengrass means vendors can take a generally available product from a leading cloud vendor and begin integrating it into their hardware.’

Not everyone agrees that edge computing will be vital to IoT. Boston CIO Jascha Franklin-Hodge argues that, ‘I think edge computing is highly overrated. There are some very specific use cases where edge computing is the antidote to not enough bandwidth and not enough connectivity. What cloud infrastructure has taught us over the last 10 years is that centralized, high-efficiency computing infrastructure in most use cases is going to outperform distributed, lower-efficiency systems in price, performance, scalability, resiliency and all the other things we value. I think a lot of the use cases of edge are going to fall off as we build more robust networks.’ Whether he’s right or not, as the IoT expands, we need to build an infrastructure that is able to handle it. Edge computing does not replace cloud computing, but with tech giants investing heavily it appears they at least disagree with Franklin Hodges, and with good reason - the combination of edge and cloud, or an as yet undiscovered replacement, will be necessary in the future, and it is worth investment.

Comments

comments powered byDisqus
Data culture small

Read next:

Building A Culture Of Data

i