Intelligent CIO Middle East Issue 52 | Page 45

IoT initiatives will change WAN requirements Many unsuccessful IoT initiatives are characterised by their failure to consider how the projects will change the requirements for Wide-Area Networks (WANs). According to Gartner, there will be 20 billion Internet connected objects by 2020. This influx of devices and sensors will significantly increase network traffic and unused to the volume of activity, traditional Multiprotocol Label Switching (MPLS) links will become congested. This will result in a decline in application performance triggering a whole host of problems, including reduced productivity and loss of revenue. In an attempt to overcome such capacity limitations, many companies will turn to alternate connectivity options, for example cellular LTE or broadband Internet. However, the routers and patchwork of tools required for this process will create an unnecessarily complex hybrid network when combined with the volume and diversification of IoT devices. www.intelligentcio.com “ FOR IOT PROJECTS TO BE SUCCESSFUL, IT MUST PLACE THEMSELVES BACK IN THE CONVERSATION. Software-Defined WAN (SD-WAN) circumvents this complexity. Unlike traditional infrastructure, SD-WAN will automatically direct traffic through the most efficient route available, regardless of connectivity type, to ensure the seamless connectivity required for a successful IoT deployment. Ultimately, this simplified, intelligent and flexible approach to network management means IT administrators can quickly deploy new services, policies and even entirely new sites, through a centralised management console. As well as being capable of addressing evolving needs. Edge Computing is vital for real- time data analysis The cloud is a key facilitator of digital strategies. Particularly when it comes to business intelligence, archiving and centralising access. However, for data- rich applications, or in scenarios where employees need to turn data into actionable insights as it’s created, Edge Computing is far more suitable due to its lower latency. Let’s take the IoT system on a manufacturing floor as an example. The system may have been designed so that the production line will shut down automatically in response to the sensors indicating that a quality threshold is being exceeded. However, if latency causes a delay in the relevant information being conveyed from the sensors, faulty units will continue to be produced, ultimately increasing downtime and overall costs. The manufacturer needs more computing power at their remote INTELLIGENTCIO 45 45