Edge computing has become a topic of hot conversation as the technology capable of supporting sensor-2-server data transport has matured. The realization of true edge computing is accompanied by a host of benefits, including real-time data transmission, maintenance needs and considerable savings for operational expenses. Is edge computing the cut-and-dry future?
Ken Briodagh, editorial director with IoT Evolution, plays devil’s advocate on a recent podcast with FreeWave Technologies CMO Scott Allen. He asks, essentially, “If companies focus resources on the real-time data transport at the edge – sending small packages of data at a time in the interest of speed – are we losing the benefits of big data? Do we lose the information that big data sets can provide in terms of predictive analytics and, ultimately, machine learning if we discard bits and pieces of data at the edge that we’ve deemed irrelevant?”
Listen to the podcast below for Allen’s response!
Overall, edge computing has three main drivers:
- latency–our need to have the data in milliseconds;
- loss of communication–able to solve the factory problem without shutting down the entire plant;
- proximity–sensors in the field monitor the data back to the edge.
Edge Computing Solution
Depending on the industry, a mixed bag of both programmable and edge computing solutions is an answer to Briodagh’s question. In some cases, especially with the oil and gas industry, companies rely on a sensor-2-server stream of communication, where they need to have the information in real-time, and if there is a problem, be able to act locally and fix the issue before anything drastic happens. The network is a combination of radios communicating with sensors that pass the data to a gateway and up to a cloud system. The network uses only small data sets to transmit a continuous flow of intelligent, sensor-based information, optimizing bandwidth in situations where latency is crucial.
Next for the Edge
There will come a time when using edge technology will just become a regular line item expense needed to do business in this modern age. Some early adopters have already started using gateway systems as a cookie cutter roll-out for all future expansions. Many worry the cost of entry is still too high to integrate, even though the need for transmission is great. As our digital age grows, infrastructure complexity and the desire to implement the latest technology grow along with it. Altogether, edge computing is still in its infancy stage, so no one really knows what data we deem irrelevant today will be vital tomorrow.