IoT and Edge Computing Services
The Internet of Things (IoT) is one of the most hyped topics in today’s technology coverage. But what really is IoT? The term IoT summarizes all connected “entities”. The Internet of Things is made up of the smallest machines up to the largest computers. All of them have in common that they use networking technology to communicate with each other. There are various forecasts of how the amount of connected devices is literally “exploding”. Intel, IDC and the United Nations published a forecast that predicts the number of connected devices to reach 200 Billion by 2020 ( https://www.intel.com/content/dam/www/public/us/en/images/iot/guide-to-iot-infographic.png ). This is a huge number, but realizing that this means 26 connected devices for every human on earth, it becomes clear how realistic this prediction is. A lot of upper income single households in the developed economies are already close to the number of 26 connected devices.
And let’s not forget that the term IoT can be seen as an umbrella term that includes the Industrial Internet of Things (IIoT). To show all the buzz around the topic IoT and the challenge to get accurate estimations IDC published estimations of 13 Billion connected things in 2015 and a predicted 30 Billion connected things in 2020 ( https://www.idc.com/infographics/IoT ). There are huge discrepancies between the published numbers of connected devices and this might reflect the rapid growth in this sector. Estimations from wireless network providers predict connection densities of up 1 million connected devices per square kilometer in their upcoming 5G networks. Knowing that all these devices collect and transmit data, we can realize what a huge growth in data this represents.
It is safe to assume that the rate in growth of data is increasing at a faster speed than bandwidth. This is important because the connected devices collect data through their sensors, but a lot of processing of this data does not happen on the device itself. IDC states that 40% of the data will be processed at the edge. So, what can we understand as the “edge”? There is the device edge, so the device itself, e.g. the handset, plus the network edge, in case of wireless networks this will be the base station, to which the device is connected. In today’s setup of networks in IoT, the data is collected on the device, e.g. by a machine in manufacturing and minimal processing takes place on the device that collects the data. Then most of the data is transferred to the core cloud, where the compute resources are located. On these resources the data is processed and insights and intelligence is generated, which can trigger further actions. That means we are talking about centralized processing on cloud-based resources.
With the increasing amount of data being collected, the question arises if we will drown in data? The following article from gives a good answer in the case of Industrial IoT (IIoT): manufacturer’s won’t drown in data if they leverage edge computing ( https://www.engineering.com/IOT/ArticleID/16715/Will-IIoT-Drown-Manufacturers-in-Data-Not-if-They-Leverage-the-Edge.aspx ). One set-up is described by Siemens, where edge computing allows large data to be processed locally ( https://www.siemens.com/press/en/pressrelease/?press=/en/pressrelease/2018/digitalfactory/pr2018040239dfen.htm ). Considering all this we can summarize that IoT is driven by edge computing capabilities. In this case, decentralized processing of data away from the cloud and closer to the source.
As we are thinking today in terms of platforms, we can conclude that, IoT is a platform itself upon which applications are built. And the successful implementation of IoT will include artificial intelligence (AI), machine learning (ML) and many more future technologies. The Gartner Hype Cycle for emerging technologies gives a good overview of the current state of these technologies and states that edge computing and IoT platforms are currently believed to be at its peak of inflated expectations ( https://www.gartner.com/smarterwithgartner/top-trends-in-the-gartner-hype-cycle-for-emerging-technologies-2017/ ).
What will it take to take these technologies to a plateau of productivity? We need to solve how do we accomplish edge computing. Will the edge computing infrastructure be widely available in the same way like connectivity or will it be available to a small amount of players and areas? In the past we have seen that economic network effects are fundamentally important for rapid development of technologies. Another issue to solve will be how we can ensure trust and security for the large amounts of data that will be passed between devices. How can security and privacy be ensured for the large amount of devices that have our data and are suddenly exposed to cyber attacks? How do we monitor and update the software for these devices?
Further Reading for examples of IoT across Industries: