Forget Cloud Computing, it’s all About the Edge

Posted on Updated on

Forget Cloud Computing, it’s all About the Edge

By: Daniel Meighen

For those of us that have smart devices in our homes which can be controlled via voice-recognition, the conveniences they provide can be taken for granted. When I initially set up the first of many “Alexa” devices in my home, I was amazed that I could do simple tasks, like turning off the lights, without even leaving the couch. But after a while, I started to take these conveniences for granted and began complaining about the lag time between voice commands and the actions that happened. Although this was not more than a second or two, it created a disconnect in the conversational aspect of using these devices. That pause while waiting for Alexa to tell me if my doors are locked is because Alexa has to communicate with the smart-hub which then queries a server on the current status of my door locks. This process involves cloud computing, where data processing happens off-site and over the internet (“What is Cloud Computing?,” n.d.).

Although there are many benefits to using cloud computing, such as low maintenance costs and scalability, there is an inherent issue of lag that occurs. Now, in most cases the lag is minimal and not much of an issue. However, I would stress the importance of reduced lag in the case where a split-second delay can be the difference between a realistic, conversational reply and an awkward, eternal pause in dead-silence while my guests are waiting to continue their conversations until Alexa decides to let me know that the oven timer has been set. This is where edge computing comes into play. Edge computing is, unlike cloud computing, where the processing of data happens locally and as close to point of data collection as possible (“What is Edge Computing?,” n.d.). For example, if I wanted to find out if my doors are locked, instead of reaching out to a remote server somewhere, my smart-hub could just search for the information locally, process the data, and send it back as a voice response. This would eliminate the transmission time of sending data over the internet and waiting for it to be processed before sending it back. Not only would this reduce bandwidth usage and processing times, it would add increased security by reducing data exposure over the internet (Ismail, 2018).

Now, what if I wanted to use this concept for analytics? Let’s say I’m a coffee addict who has a smart coffee maker that makes the perfect brew at the command of my voice, and with the new year coming up I figure I could make a New Year’s resolution to cut back on my caffeine intake for a better and healthier me (who am I kidding, right?). Through edge analytics, the coffee maker could process my daily “brews” and create an ongoing analysis of how I should be cutting back to reduce my average intake of caffeine. This could all happen locally without needing to send the data to the cloud for processing and analytics. This is what separates edge computing and edge analytics. Edge analytics pushes edge computing further by taking the locally processed data and turning it into usable information on the device (Ismail, 2018).

So what about scalability? Although I’m all for healthy living, I don’t see a purpose in analyzing the coffee data from the hundreds of coffee makers in my home. However, we can look at this from the perspective of manufacturers. Edge analytics would allow large pieces of machinery to process data from their sensors to perform long-term analysis and act on them in a predictive manner. For example, a machine which packages tissue paper may have a part that comes out of alignment periodically which cause the machine to bind up on the tissue paper. With edge analytics, the machine could collect data on past failures to create a predictive analysis on when the next bind-up is likely to occur. The machine could then self-align in anticipation of a possible failure and prevent the event from ever happening. By not needing to send the data over a network and wait for external processing, the machine could make adjustments in a fraction of a second based on an analysis of the data internally. This capability could save companies millions of dollars in unnecessary and wasteful product issues. And who knows, maybe this will also mean I won’t need to deal with those awkward pauses from Alexa anymore…


Ismail, K. (2018). What is Edge Analytics? Retrieved from

What is Cloud Computing? (n.d.). Retrieved from

What is Edge Computing? (n.d.). Retrieved from