As 5G networks prepare for launch in early 2019, 5G technology and Fog Computing are the intelligent network architecture of the (near) future.
With official rollout dates just around the corner, it’s finally time to get excited about 5G. Across the globe, deployments are already set to begin in 2019 in the US, China, UK, and other major markets.
However, the real value of 5G may well not reside with familiar consumer devices. One particularly compelling use case for 5G is in combination with Edge AI , a burgeoning market that is estimated to reach $51.6 billion by 2025, in a new network layer architecture: Fog.
Visions in the Fog
The increased data capacity of 5G networks over existing 4G options should bring high-speed data connections to remote areas, as well as enable far greater numbers of IoT devices to seamlessly connect without unnecessary contention. And while 5G data speeds will provide low latency backhaul to the current cloud architecture, there are significant benefits to Edge AI that are accentuated with 5G speeds.
Fog Computing is essentially a middle layer of intelligence baked into the data network, where input from remote, distributed IoT networks is processed anywhere at the edge of the network, rather than being phoned home.
This not only takes advantage of 5G’s ultra-low latency and high bandwidth, but also enables operators to maintain a flatter core network architecture that will improve “base station-to-cloud” latency, as well as squeeze core network rollout costs.
Automotive applications – latency is key
One of many key applications for 5G is predicted to be smart or connected cars. As Machina Research recently forecasted, “IoT will account for one-quarter of the global 41 million 5G connections in 2024” – with 75% of these in the auto industry via embedded vehicle connections. Latency in the world of smart cars is particularly important – a connected car traveling at 75 miles per hour would travel over 10 feet further before applying the brakes if the system experienced a mere 100-millisecond delay.
Combating such delays and latency in automotive applications is where onboard edge AI systems can be of critical value, especially in preventing an emergency situation. The new VIA Mobile360 M820 Edge AI system was demoed recently at the China International Industry Fair held in Shanghai. It powered an in-vehicle cockpit integrating ADAS, 360-degree and surround view capabilites, which also included a driver behavior monitoring application that scans for common signs of fatigue or inattention in the driver.
Enterprise Edge AI meets Fog
Of course, the potential of Fog Computing in 5G-enabled Edge AI environments extends way beyond the automotive industry, bringing the benefits of AI at the edge to all manner of enterprise and vertical markets. Baking intelligence into the network, closer to original data sources, will deliver significant benefits and drive innovation. For example, processing ML (Machine Learning) and deep learning (DL) model data at origin, and only sending back the very highest quality data for training purposes will lead to significant improvements in ML performance, as well as enhancing overall security, and reducing backhaul costs.
Smarter cities emerge
Meanwhile, innovations such as smart cities and smart utility grids will also benefit significantly from Fog technology. Utility grids in particular, where data from remote sensors can be processed and then acted upon far more rapidly than the current cloud-based models, will be able to increase everyday efficiency, as well as respond to demand fluctuations far more dynamically. The big value that 5G brings to the table in this scenario is the large number of concurrent devices that 5G cells support compared to 4G. This means that individual components of the smart city or utility supply chain can be connected in their own right, sharing data and also – with edge AI technology – aggregating and actioning that data in real time.
Overall, while 5G will continue to hit the headlines for blazing data speeds, the real news is the way in which it will combine with edge AI to fundamentally change the way tomorrow’s mobile networks will handle and process data. The results will enable IoT expansion at an unprecedented rate, delivering truly intelligent, scalable networks to meet future demand.