With the explosion of IoT, connected devices are collecting more and more information through sensors, cameras, accelerometers, LiDAR, and depth sensors. IoT use cases are extensive and encompass everything from connected cars, factory machinery, agricultural processing, wind farms, oil and gas refineries, wearable technology, smart home devices and just about every piece of connected equipment you could imagine. This article takes a look at the challenges caused by this data deluge and the solutions provided through edge computing in three use cases.
More devices, more data, more problems
With the greater number of connected devices in operation, the sheer volume of data being collected is astronomical. Capturing, aggregating, and analyzing data becomes a greater challenge. Not all data is useful, yet some time sensitive data such as autonomous vehicles, noxious gas monitoring, healthcare, safety equipment, and other scenarios are at risk. A split-second delay of data going to the cloud and back to the device could be disastrous or deadly. Other data sites face the challenge of location where the use of IoT in rugged environments such as an offshore oil refinery, underground mine or deepwater well can result in unstable links with limited bandwidth and variable latency.
The solution is edge computing - or an edge/cloud hybrid model where time-sensitive data is processed at the edge and less urgent analytics can be determined in the cloud. Edge changes possibilities and enables new scenarios that are/were not possible/effective with cloud processing. Connected Industry is paying attention, and a 2015 report by IDC forecasts that by 2019 45% of IoT created data will be stored, processed, analyzed and acted upon close to or at the edge of the network.
Scalability and cost reduction in wind farming
Edge computing is fundamentally ‘distributed computing', meaning it improves their saliency, reduces network load, and is easier to scale. Data transmission costs are lower because the amount of data transferred back to a central location for storage is reduced. It follows that this facilitates a much more autonomous and decentralized model that creates a real opportunity to create more customer value.
An example is the transmission of wind farm data. Wind farms can consist of hundreds of wind-powered turbines generating vast amounts of data. According to researchers, Wikibon, a typical wind farm was embedded with security cameras and other sensors with a distance of 200 miles between the wind farm and the cloud. Through processing data and the edge transmitting summary data to the cloud, they reduced traffic flow by 95% and reduced the cost of management and processing to $29,000 from $81,000 over three years.
Real-time data insights
As connected cars evolve from level 2-5 in automation, the need for real-time data is becoming more critical. A moving vehicle could ostensibly be considered a computer on wheels. Autonomous vehicles will hold terrabytes of data that is critical in enabling the vehicle to make split-second decisions such as detecting obstacles and navigating all different kinds of terrain. As we'll be sitting on level three for some time, companies are working to connect the experience of the driver and the vehicle, again requiring real-time data.
At CES this year, Nissan unveiled research that will enable vehicles to interpret signals from the driver’s brain, redefining how people interact with their cars. In the world’s first system of its kind, the driver wears a device that measures brain wave activity which is then analyzed by autonomous systems. By anticipating the intended movement, the systems can take actions – such as turning the steering wheel or slowing the car 0.2 to 0.5 seconds faster than the driver while remaining largely imperceptible. The company’s Brain-to-Vehicle (B2V) technology promises to speed up reaction times for drivers ultimately making driving more enjoyable and in the long run, safer. The data generated could also have monetization value in selling driver data to car manufacturers, insurers and health researchers, and could lead to a greater understanding of the capabilities of the brain-machine interface in other applications.
Predictive Maintenance & Monitoring
A key component of digital transformation is updating (or in some instances replacing) legacy equipment to produce digital monitoring to benefit a company. Litmus Automation worked with a commercial industrial boiler manufacturer whose assets include a long lifespan of forty to fifty years. By digitizing these machines, a new revenue stream and business model was created as a result of understanding the company’s asset data. This not only supplied them with better analytics of their own assets, but also provided value-added services for their consumers and users. This allowed better visualizations and real-time machine health monitoring to enable them to predict machine faults before they occurred. This real time health monitoring was made possible through the use of edge computing to provide crucial data insights. This solution not only created a value to the end user by decreasing failure costs, but also helped generate brand awareness and customer loyalty.
With the growth of emerging technologies such as AR, VR, AI, cognitive computing, and blockchain technologies, we can expect to find the prevalence of edge computing growing and expanding across Connected Industry. Innovation shows no sign of abating and the challenges of the data deluge will remain with adoption of edge computing. It's a technological advancement set to bring great benefit to Connected Industry.
Interested in finding out more about how Edge computing can help your business? Take a look at our webinar "The Intelligent Edge" and sign up to receive our upcoming white-paper.