Edge Computing Enables Intelligent Networks

Edge computing is the key factor to make the Internet of Things work since connected devices will generate so much data that transmitting, storing and analysing all that data at a central location is no longer viable. Not only that, connected devices such as drones, self-driving cars or robots will, most likely, require extreme rapid processing. Creating the data, sending it to the cloud for analysis and returning the results to the device will simply take up too much time.
The predictions are that in the coming decade, we will add approximately 100 trillion sensors to our global economy, generating an unfathomable amount of data. The solution for all this data that requires rapid processing is doing edge computing; computations on the sensor itself, albeit at first this will be done on the device instead of on the sensor. Peter Levine, a general partner at venture capital firm Andreessen Horowitz, even believes that edge computing will slowly take over cloud computing. Although that might sound pretty crazy, it also seems very logical. Today, an average self-driving car produces approximately 1 Gigabyte of data per second, which will likely increase in the years to come. Having to send that data to the cloud, analyse it and return the results would simply not work.
Therefore, in 2018, we will see increased attention to edge computing to enable intelligent networks, where connected devices will perform the required analytics at location and use the results to perform a certain action. It will happen in a few milliseconds, instead of the few hundred milliseconds it takes today when using cloud computing. With self-driving cars that difference can be the difference between a crash or a safe ride home. The world’s cloud computing giants are not ignorant about the opportunities of edge computing. Microsoft has developed Azure IoT Edge and Amazon recently developed AWS Greengrass. In addition, startups such as Packet and Vapor IO are also bringing cloud computing to the edges. In 2018, edge computing will find its way to connected devices, before truly taking off in 2019.