What is edge computing and how is it different from the cloud? Find out in our latest Tech Explained feature where we explain the practice of processing data in real-time without delay.
At Springwise, we are covering the continuing evolution of smart devices – devices like smart mirrors, connected air vents and interactive beer labels. Today, most of these devices use the cloud to process the data they receive. In the future, however, these IoT devices will likely include the ability to process much of this data autonomously. Welcome to edge, or fog computing. Edge computing involves shifting processing power from the cloud to devices like smart refrigerators and automated manufacturing systems. So, what exactly is edge computing, and how does it work?
Edge computing is the next generation from cloud computing. In cloud computing, data is moved away from individual or company computers and into data centres run by tech companies like Microsoft and Google. Dropbox, Gmail and Slack are examples of cloud-based applications. However, more and more devices are now relying on artificial intelligence. AI requires a lot of data that often has to be processed in real time, making it harder to depend on the cloud. As a result, there has been a trend to move critical processing and decision-making into the devices themselves. These ‘edge’ devices are connected to the cloud, but they also have enough onboard computing power to ‘make decisions’ on their own.
Edge computing is about moving processing and storage closer to the application or the end user. For example, smart thermostats like Nest adjust the temperature in users’ homes based on the data they collect about owners’ habits. Devices like these are edge devices – they are able to adjust the temperature in real time without referring back to a cloud-based algorithm for instructions.
One big driver of edge computing will be autonomous vehicles. Self-driving cars generate as much as a gigabyte of data every second, and even a microsecond delay in processing this information could potentially lead to an accident. These vehicles will need to be able to crunch data without first uploading it to servers and waiting for a response. This is why developers of self-driving technology have been equipping their vehicles with heavy-duty processors. They are giving each car enough computing power to become its own data centre.
All of this sounds a lot like a return to the days before the cloud, when computers were more autonomous. Yet, the goal of edge computing is not to replace the cloud, but to work in concert with cloud-based platforms. For example, autonomous vehicles will not be completely independent of the cloud. They will send the data they collect to the cloud, where manufacturers can then use the data to further refine their software.
Edge devices will do most of the initial processing, but rely on cloud intelligence to improve their functions. This is why all of the major players in cloud computing are also embracing edge computing. Some are also working on making websites and apps more edge-like. For example, Google is developing apps that have offline-first functionality. Users will be able open a app offline on their phone, save any changes locally, and then sync up with the cloud later to upload those changes.
One big advantage of edge computing is the potential bandwidth savings. For example, instead of uploading all the data to the cloud, edge devices could choose which information to upload. The other potential advantage to edge computing is greater security in IoT. There have been a number of issues recently with IoT devices that can be easily hacked. Adding processing power to a smart toaster, and then managing security centrally from the cloud, could make it much harder to hack the toaster.
15th January 2019