From ultra-connected autonomous cars to low-latency AR, VR and gaming: to remain competitive in the digital age, businesses will have little choice but to fully embrace the new opportunities that come with the deployment of edge computing, according to a new report published by Intel.
Almost three-quarters (72%) of IT leaders are already using edge computing to provide innovative services, according to the chip giant, whether that is to create new products, open new revenue streams or boost efficiencies.
"Businesses can no longer afford to ignore the edge," says the report, stressing the technology's potential to better access and understand the unprecedented amounts of data that are generated over networks every second.
As the name suggests, edge technologies consist of moving the hosting of computer services to the edge of the network, so that the process happens as close as possible to the people that use the service, which significantly reduces latency.
Edge computing, therefore, happens in devices distributed across the network, which is why Intel refers to a new age of "distributed intelligence", in which any and every object can be made "smart" and connected, so that it can store and process data to enhance the delivery of a given service.
This could be useful in a number of fields. In healthcare, for example, edge devices could take the shape of integrated wearables that track patient vitals throughout the day and transmit data for ongoing evaluation and treatment. The technology could also be deployed in autonomous robots to support health workers; or in image-based diagnostics, to power deep-learning models and speed up the detection of health issues.
The core reason for deploying edge computing is a trend shared by virtually all industries: the explosion of data. As users increase their reliance on digital technologies, so is the amount of data they generate reaching unprecedented levels. All of this information can be processed and analyzed by businesses to improve their services – but it requires huge computational power.
Sending all the data that is created back to the cloud for processing is impractical and causes latency issues. Enter edge computing: by placing compute nearer to the source of data, whether it is a smartphone, a PC, an IoT device or a sensor, the volume of information that is generated becomes a lot easier to handle.
AI scientist Inma Martinez, who led the research for the report, said: "The edge makes possible a world where all of a sudden, every single object has the potentiality for information – information that can be extracted and used in real time."
In turn, this lets businesses innovate in a number of different ways. In retail, for example, in-store digital signs and interactive kiosks can help gather a better understanding of customers' expectations and behaviors. Smart sensors and cameras coupled with AI-powered analytics can respond to touchless gestures and inform retailers on the effectiveness of a message or a display.
The technology can also help stores provide personalized services: Intel reports that Italian luxury computer vision boutique Wonderstore, for instance, uses visual sensors powered by edge technologies to analyze customers' behaviors and customize their store experience based on data related to their dwell time, for example.
In industry, edge computing is also expected to revolutionize the working of the factory floor. The technology will enable the real-time processing of raw data generated by machines to improve predictive maintenance or check for defects.
Similarly, edge technologies can power AI-based robotics that are used in manufacturing to perform repetitive and dangerous tasks. Car manufacturer Audi, reported Intel, has seen a 100-times boost in its quality control check times since starting to deploy computing at the edge, with only 18 milliseconds of latency – which has in turn reduced labor costs by up to 50% at the firm's Neckarsulm site in Germany.
Edge computing, therefore, comes with the ambitious promise of significantly reducing latency in all sorts of applications and use-cases, which in turn will unlock innovative services in many different industries.
Still, despite Intel's enthusiasm, it's perhaps not entirely surprising that businesses are not focused on edge computing yet. For all the excitement, the technology remains immature, and viable business models making use of connected devices and internet of things devices at the edge of the network are still in the making.
This is in part due to technical challenges: for example, edge computing is highly reliant on faster connectivity – which is why it is often portrayed as going hand-in-hand with 5G connectivity. And although 5G roll-outs are gathering pace, next-generation connectivity is still far from widely available to all.