Many industries will expand their reliance on IoT devices to increase operational efficiency, transform customer experiences and drive growth. But IoT has limitations and challenges when it comes to data storage and processing, many of which can’t be solved using the traditional public cloud model alone.
The solution? Bring the cloud to IoT with edge computing. Edge architecture is computing that occurs at or near the source of data, providing processing “at the edge” before data makes its way to the cloud.
Edge computing will enable massive deployments of IoT devices. But it’s not all about scale. Edge will transform what IoT is capable of in in several fundamental ways:
- Latency: By including computing in an IoT device that would otherwise have to communicate with the cloud to deliver a response, latency is greatly reduced. An example of this is the rumored upcoming inclusion of AI chips in Amazon’s Alexa.¹ This would let Alexa provide results faster, as well as reduce Amazon’s cloud server costs.
- Bandwidth: Currently, IoT devices have a bandwidth problem. For instance, an organization with one security camera can stream all of its footage to the cloud. Attempt to stream all footage from twelve cameras, and there is a problem. Edge has the potential to change this. By including smart edge technology, a security system can potentially upload only “important” footage, as defined by preset rules.
- Privacy: IoT devices are notoriously susceptible to cyber-attack. By attaching edge computers to smart devices, security measures such as encryption and identity management can be implemented for IoT. Smartphones are already doing this. Azure Sphere, a managed Linux OS, is a certified microcontroller that is designed to make IoT devices as hard to hack as a standard PC.
The following three use cases are examples of how edge computing has the potential to transform IT infrastructure and drive business outcomes.
Of every sector, industrial manufacturing may have the most to gain from IoT and edge computing. The incorporation of storage and computing into manufacturing equipment will enable ongoing data collection and analysis. This lets organizations customize production to meet demand, predict maintenance requirements, and increase energy efficiency.
Edge will be a primary factor in the success of IoT for manufacturing. This is particularly true of operations with low latency and/or bandwidth. For example, offshore oil rigs can leverage edge computing to gather and process environmental data without relying on a cloud data center infrastructure. This enables managers to make split second decisions based on current data.
Edge technology has already been adopted by many financial institutions. Banks are including edge into ATMs and kiosks to gather data for faster response times and a more robust set of features.
Perhaps the most significant impact of edge for financial institutions is seen in hedge fund and finance firms dealing in the stock market, where a millisecond of lag in a trading algorithm can result in a substantial loss of money. Edge computing lets these companies place their servers as close to stock exchange data centers as possible, letting them run resource-intensive algorithms right at the source while ensuring they have fast access to the most current financial data.
For healthcare, IoT and edge represents an exciting opportunity for exploring new ways to provide patient care. Edge and IoT devices are already being used to deliver massive amounts of patient-generated health data (PGHD), which provides healthcare practitioners with the access to critical information in real-time, as opposed to receiving incomplete data from a relatively slow database.
Edge computing could also have a significant impact on the delivery of healthcare in remote, underserved areas where professionals often lack access to critical medical records. With edge-enabled devices, data can be stored and processed on-site, with information delivered in real-time. Medical devices with edge technology may even be able to recommend treatments.
¹Amazon may be including AI chips for Alexa