It’s funny how the pendulum swings in the technology world. First we started with mainframes, then moved to personal computers, now we’re back to the cloud. Remote, local, remote. Right now we’re on the forefront of the pendulum swinging back again, this time to edge computing.
Today’s blog post is all about possibilities. The what if? What if we could push computations back to the edge? What if data was processed immediately on ingest? What if we didn’t need the cloud for basic processing? What if you didn’t need Internet connectivity to get your data in real-time?
Think back to those mainframe days. Or in my case, think back to when you read about those mainframe days in a textbook. You had a single centralized computer (mainframe) with hundreds of dumb terminals that connected to it. These terminals didn’t have any onboard processing, they only served as an I/O device for the central mainframe. Today we’ve completely replicated this model on the edge, only instead of a terminal for human I/O we have the dumb sensors that simply read and stream data to the cloud. The difference though is that these dumb sensors aren’t actually dumb. In fact, they have plenty of untapped potential.
In most edge devices, processors sit idle, only waking to read a sensor and push the data to some other device. This is a total and complete waste of processing power. Most of today’s edge devices are more powerful than my first computer and that thing could certainly do more than a simple sensor read. So why not transition processing to these devices? Honestly, I’m not sure. Maybe it’s a reliance on the cloud, but there has to be a way to capitalize on these already existing resources.
I’ve done a lot of thinking about this and I can’t figure out why the transition hasn’t already happened. I thought for a while that it could be due to a lack of edge database offerings, but SQLite has been around a while. Quick plug, checkout the HarperDB vs SQLite benchmark. Perhaps it’s the added complexity of storing data in multiple places or maybe we’re just intimidated by the intelligent edge. There’s an illusion of added complexity to having edge computing as well as cloud computing running together. However, I’m not sure it’s actually more complex. In fact, I think it has the potential to be much simpler.
What if we did our initial processing on the edge? Real-time number crunching on the edge, which if it’s based on a single sensor, is not terribly difficult or complex. Now our intelligent edge device has filtered out all the data we don’t need, storing only the important, useful data in the edge database. Our data footprint is trimmed down immensely, which also means the network load is reduced. This is a huge benefit for many IoT cases where the issue isn’t power, but network connectivity. What about real-time response rates? Many common edge computing solutions today retrieve data from the edge and send it to the cloud for storage and processing. If we then need to retrieve this data in real-time back on the edge we have to go up to the cloud to get it. Does that make any sense? Let’s take a farm for example. We could have IoT sensors out in the field (literally a field for once) reading data with a dashboard in a tractor displaying real-time data to the farmer. We could either have tons of 4G cards reading and writing data to the cloud or we could setup a local network with edge database and never bother to have the data leave the farm. The cost savings on connectivity alone should illustrate the point. It would also be significantly faster to query the devices around you than to bother going up and down, to and from the cloud all the time.
Is the pendulum already swinging back to the edge? I think so. When we were at IoT World, many people that I spoke with, wanted to hear our perspective on how they could stop relying on the cloud for everything. Our answer is simple, your data is your data and it should reside where you need it. We believe in implementing and utilizing an edge database to keep your data in your hands. Companies are starting to see the light of the edge. It’s going to be exciting watching the edge grow. Personally, I think it’ll be similar to how we watched the cloud materialize. In the beginning no one really knew what cloud computing meant, everything was the cloud. I think we’re going to see the same thing with edge computing. We get to define what the edge is and how powerful it can be. It’s going to be a fun ride and I for one cannot wait to see how powerful the intelligent edge will become.
HarperDB is cutting edge, edge technology (pun absolutely intended). Check out the product to see how HarperDB can make your edge a little more intelligent.