Return to overview

Edge Computing & the Internet of Things

Edge Computing Balances Local And Cloud Processing

Edge computing is computing that happens at the “edge” of the cloud; that is to say as close to the user as possible. Instead of the data being sent to cloud servers or a data center, it is handled locally by an edge gateway. The gateway takes the data from the device, sends on what needs to be sent, and sends back data for real-time application needs.

The edge gateway can be a laptop, a smartphone, or another device on the IoT network. This architecture keeps most of the IoT processing within the local network, avoiding latency and protecting it behind a firewall.

In other words, edge computing balances local and cloud processing; it reduces the need for individual devices to process their own data without sending everything to the cloud.

What are the Benefits of Edge Computing?

The primary benefit of edge computing is speed. By reducing latency, edge computing increases the speed at which devices respond to their users. For example, recent Nest cameras can handle most of their image processing on-device. This means they aren’t sending raw data to the cloud, waiting for it to be processed, then receiving the final version back. Alerts are faster and false positives become a lot fewer.

These data round trips might seem insignificant but can result in significant slowdowns. A smart speaker that is processing your voice itself or through a device on the local net will respond just that tiny bit faster.

There are several other benefits, however:

1. It can still work if connection is lost

Devices will continue to work, at least in a limited way, if connection to the wider internet is lost. For example, a typical cloud-based smart speaker will simply stop responding if it loses its connection to the cloud server. Any local processing, however, can continue. For critical systems that can’t afford downtime, this can be vital.

2. Improved privacy and compliance

Data that is not sent to the cloud cannot be intercepted; data stored in multiple places is less likely to be copied or lost. However, it is important that the individual devices are properly secured. Data within the network may need to be encrypted and some systems use VPN style tunneling.

3. DDoS resistance

Cloud servers often experience distributed denial of service attacks. These attacks deny access to the cloud server by intentionally overwhelming bandwidth. They are getting worse over time. Additionally, better firewalls prevent your own devices from being used as part of a DDoS botnet.

4. Reduced costs

By storing most data locally, you don’t have to pay for as much bandwidth or cloud storage. This is an advantage for users and also for device vendors who might otherwise have to provide the cloud storage to their usage.

5. Scalability

You can add devices freely without worrying about what it will do to your storage costs, as long as you have the local network set up the way it needs to be. It is often cheaper to add another edge gateway than to increase your storage on a cloud server.

What are the Best Uses for Edge Computing in the IoT?

So, what uses and devices is edge computing most valuable for? We’ve already mentioned security cameras, which handle a lot of data and need to process it. A security camera that processes locally is not going to have to wait for data from the cloud. Nor is it going to use up cloud storage on three hours of nothing in the middle of the night. However, there are some other uses:

1. Smart speakers

We expect smart speakers and voice-activated remotes to respond quickly and accurately. At the same time, many people are worried about privacy issues with what their smart speaker may be storing. If only necessary commands such as online shopping go to the cloud (without financial information), and if non-commands are deleted at the source, then the privacy concerns become much less significant.

2. Self-driving cars

Self-driving cars have to have extremely fast reaction times, ideally faster than humans. They also need to be able to operate where there is no internet and communicate with each other efficiently. It’s impractical to use cloud-based computing for self-driving cars except possibly for downloading maps from areas the car is less familiar with.

3. Health monitors

Health monitors need to be able to send the appropriate alert to patients or caregivers right away, accurately, and with no latency.

4. Smart city automation

Smart city sensors, smart traffic lights, etc, are highly dependent on the cloud, but having every traffic light go out in downtown because of an internet failure is obviously not a desirable situation. By using some edge computing, traffic signals can continue to operate and streetlights can communicate with each other without going to the central cloud.

That’s just a few of the possibilities. In many ways, the cloud makes a lot of sense, with its redundancy and backups. However, it also adds latency and security concerns and the future of the internet of things is going to involve more and more data being processed locally, more and more decisions being made by devices and local networks without consulting a central server.

Conclusion

If you are working on the internet of things applications for your company, you need to consider how edge computing can benefit you and plan your network and platform accordingly. Get in touch with the experts at Davra to find out more about how edge computing and IoT can help your organization.