Cloud edge computing
Clouds are moving to the edge of the network. That may sound like an oxymoron since the cloud is basically a centralised facility where IoT data transmitted from the edge is aggregated and processed into real-time information.
The data — Big Data — can also be analysed in the Cloud in order to find correlations between events that at first sight would not appear to be related. It’s a proven process so why change it? says, Bob Emmerson, independent M2M writer and analyst.
One reason is Moore’s Law, which states that overall processing power for computers will double every two years. Therefore, it has become technically feasible and cost effective to process data locally, close to the source, in a network of edge-based clouds. The primary reason for migrating to edge cloud processing comes from the ability to employ local augmented intelligence and in-memory computing platforms. This enables local IoT data to be analysed in real time, which in turn allows decisions to be made “in the moment”.
Edge-based clouds
Asset gateways that aggregate sensor and machine data, which is transmitted over long- and short-range wireless networks, are de facto edge-based clouds. They can be distributed within the network in order to boost network capacity and enable multiple local solutions. This development is facilitated by LoRa technology, which enables the creation of low power, wide area networks (LPWANs). Sensor nodes, which can be 2-5 km away in dense urban environments, 15 km in suburban, transmit data to a LoRa gateway, which can be an off-the-shelf product. The use of unlicensed spread spectrum technology prevents communications with different data rates interfering with each other and that enables the creation of “virtual” channels, thereby increasing the capacity of the gateway. However, LPWANs should not be seen as a substitute for long range, high capacity cellular systems.
Summary and conclusions
Forget the hype about tens of billions of devices being connected to the Net by 2020. They are over optimistic predictions that are potentially damaging to the industry and they are not relevant when it comes to the creation and implementation of enterprise IoT solutions, for which there is a compelling business case. They have the proven ability to deliver tangible benefits, but enterprises that fail to realise those benefits in time could find themselves falling behind the IoT curve.
That may sound somewhat alarmist, but consider this hard-hitting statement: “Enterprises are in many ways the laggards of the new digital era. Their processes and internal and external service remain rooted in the analog or physical world, and the digital systems they use are frequently disjointed and out of date.” It comes from a recently published book by Bell Labs titled “The Future X Network” and it indicates the need for enterprises to make a comprehensive digital transformation.
Enterprise IoT is set to play a pivotal role in that transformation process. After all, it brings analog “things” into the digital era.
The author of this blog is the independent M2M writer and analyst, Bob Emmerson.
He can be contacted at: bob.emmerson@melisgs.nl
Comment on this article below or via Twitter: @IoTNow_ OR @jcIoTnow