One industry leader – American giant AT & T recently committed to edge computing services that will see the company put micro data centres in its central offices and cell towers throughout North America.
While acknowledging that hyper scale cloud has advantages in scale and efficiency – AT & T recognises that some things need to have processing power much closer to the source of the data. The advantages will be lower latency, highly parallel, near real-time workloads that will play a huge role in many other emerging applications including IoT, Smart City applications, Smart Phone applications, autonomous cars, AR/AV and robotics.
In a recent US Government Technology publication – scientists predict that IoT and Smart City projects will force the move of more computing to the networks edge. They say that moving processing power to the edge of networks would solve many of the toughest problems associated with robotics and computing Infrastructure.
IoT, Smart City applications and data analytics mean connecting pools of data that never needed to be brought together before. Traditionally, organisations have maintained data silos successfully, without sharing – but the advent of Smart City and IoT technologies is forcing the move away from silos to more open and mutual sharing of systems and data.
Another interesting by-product of this development will see greater use of better and more powerful sensors including RFID – by adding processing capability that will allow them to act on a real time basis. The increased deployment of sensors on roads, bridges, and other infrastructure is already happening particularly in Smart City applications around the world. In years to come the current sensors will improve exponentially in terms of power, capability and intelligence.
In Smart City situations, the growing need by city leaders for more operational intelligence and a greater focus on the needs of the community will drive the need for edge computing power. Currently there are advanced IoT and Smart City applications (using edge computing methodologies) underway in Mesa, Arizona / Palo Alto, California, Chicago, Illinois, and Los Angeles. In the words of the team involved “ we are moving the algorithm to the data and not the data to the algorithm.”