Buzzword Bingo: The Empowered Edge
Buzzwords are everywhere in the electronics industry. In this ongoing series, we break them down to separate reality from marketing hype.
Buzzwords in the electronics industry usually fall into one of two categories for me. The first of which is the oh yeah, that makes sense column. The second is the wait…what? column, which usually requires more explanation.
My hope is to cover both kinds of buzzwords in this ongoing series, but this time we’re looking at edge computing, or “the empowered edge.” At its core, this concept makes a lot of sense. The question is whether or not it has matured enough to be worth your time and investment? Let’s find out.
The Elevator Pitch
Eric is the CEO of an autonomous vehicle manufacturer and he needs more processing power in the vehicle to monitor and react to changes in its environment like temperature, weather, and so on.
He can’t exactly fit an entire server into the vehicle (though he could if this hypothetical car was a Cybertruck), so instead, he relies on powerful processors that have become incredibly small while still retaining their power.
By installing them in the vehicle, in the same vicinity as the sensors gathering data, he places them at the “edge,” and removes the need for external processing or cloud computing.
An Overview of Edge Computing and What it Offers
Earlier in 2019, I reported on edge computing for the Findchips Blog and learned a lot about its role in industrial automation. Since then, I’ve learned a fair bit more about the technology and other use cases as well.
At a high level, edge computing is the concept of bringing data processing and computing hardware to the edge or source of said data. To use a term from journalism, it’s like skipping out on external leads and going straight to the source, in this case, the sensors or hardware that are gathering the data.
This means relying less on cloud computing or data transmission, and instead handling some of the heavy lifting onsite as opposed to sending it outward to the cloud for processing.
The benefits here are certainly compelling:
- Reduced Latency
- Stronger security
- Enhanced bandwidth
Bringing your hardware closer to home makes a lot of sense, there’s no disputing that.
Reality, as always, is never quite that simple.
Applying Edge Concepts to The Real World
The concept of edge computing is something that a lot of industries can benefit from. While a factory may utilize it for monitoring machinery and making adjustments for efficiency, a car manufacturer may use it to put more computing power into a vehicle that has autonomous elements.
It doesn’t stop there. Having the right hardware handy can make surveillance and facial recognition happen faster than ever without the need to process the images at a remote data center.
As it currently stands, edge computing is the most at home in industrial IoT applications where data is collected and analyzed. The benefits for this industry are closest to what I listed above.
All of this comes at a price. These are the challenges we still need to overcome:
- Hardware limitations: We need to consider the computing power available to us in smaller form factors. You cannot expect the same level of hardware as a data center in something a fraction of the size.
- Monitoring and Maintenance: Any edge computing solutions must also offer remote management without the need for physical intervention. Since these devices will not be centrally located, they need to function independently from a maintenance standpoint.
- Data Handling and Backup: While edge devices offer their own security benefits, they are not immune from issues. Specifically, you will need to keep data secure as it's transferred to larger data centers, and ensure that it’s secure on the local device.
The Final Verdict
Edge computing is more than just a buzzword that will fade with time. While it has begun to proliferate in some respects, the true catalyst for this technology will most likely come in the form of 5G rollout.
A report from Allied Market Research predicts that the edge computing market will reach $16.55 Billion by 2025, and part of that growth is attributed to the benefits received from 5G’s predicted speeds and bandwidth.
Improved capabilities could also lead to more industries adopting hardware at the edge. Autonomous vehicles seem like a safe bet, where local processing could adapt to the ever-shifting environments that a vehicle would face during travel.
A partnership with 5G will help edge computing reach critical mass. Until then, this is one buzzword to keep an eye on, but if you plan on implementing your own solutions, keep the current limitations in mind.
