The Internet of Things (IoT) is stretching beyond our bodies, homes, and vehicles to include things never-before networked: equipment, machines, sensors, and more. As the industrial IoT (IIoT) expands, connecting everything everywhere, a continuous stream of data is being generated — a whole lot of it. Cisco predicts that there will be more than 3X as much cloud traffic in 2020 (14.1ZB), compared to 2015 (2.9ZB). Because some of that data can’t — and shouldn’t — be sent to the cloud, manufacturers are looking to edge processing for alternative solutions. And that’s where Qualcomm Technologies comes in.
Edge processing defined
Processing data "at the edge" means processing happens closer to the source — via a gateway device, for example — instead of traveling to the cloud. Today, the term “gateway” can mean a lot of different things. At Qualcomm Technologies, we define it as a device that handles southbound communication (collecting raw data or information from sensors and devices), the processing (taking that data and processing it locally at the gateway level), and northbound communication (communicating this information to a backend cloud).
Benefits of edge processing in gateway devices
Lower costs: Sending all of your data to the cloud for analysis, processing, and storage can be cost prohibitive. Video in particular gets expensive quickly. With edge processing, only necessary data is sent to the cloud. The rest is handled locally, thereby reducing costs.
Lower latency: There are certain control applications that are served better by local processing, especially when time is of the essence. Let’s say there’s a piece of equipment in a factory that’s running dangerously hot. You want to be able to adjust that sensor immediately, so intelligence at the device level is key.
Increased privacy: Data protection becomes even more important as billions of sensors and devices connect within the IIoT. Many companies are handling data that they don’t want, or that can’t, be sent to the cloud. That data can be processed at the gateway instead.
The advantages of edge processing
Add in Qualcomm Technologies’ wireless expertise, and this is where it gets really exciting and advantageous for manufacturers who are developing IoT gateways.
Let’s face it: Wireless can be difficult. But not for Qualcomm Technologies. And edge processing requires various classes of communication technologies — short (Bluetooth, 15.4, PANs), long (Wi-Fi, LANs), and wide range (WANs) — to work together. These are all standards-based technologies that have been developed independently. We’ve spent a lot time in R&D identifying ways to get them to work together without negatively impacting each other.
We have proven solutions out in the ecosystem, and we continue to work with standards organizations, like 3GPP, to ensure reliability and building to specifications. So, when it comes to IoT solutions for communication and processing, it makes sense to work with someone like Qualcomm Technologies that can offer standards-based, reliable, and proven LTE IoT technologies.
This includes security, which is of paramount importance to anything dealing with the IoT. We’re placing security at the hardware, or chip, level. Our field-tested security solutions have shipped in billions of devices over the years. And we’re bringing that same level of Qualcomm Snapdragon class security and success to solutions for the IoT.
Qualcomm Technologies expertise at work
We're working with companies to develop solutions that make it easier for OEMs to deploy IoT gateways. Our combined goal is to improve operations, reduce time to market, and help ensure success.
To that end, we recently announced support for Amazon Web Services (AWS) Greengrass in select SoCs, including the Snapdragon 410E. By combining the processing power of the processor’s CPU, GPU, and DSP with the AWS software, compute, messaging, data caching, and sync capabilities can happen at a local level. Manufacturers expected to commercialize solutions running AWS Greengrass on Qualcomm Technologies’ chips include Advantech and Thundersoft, with commercial products expected to be available later this year.
We’re also working together with Sierra Wireless on its feature-rich, low-power mangOH Red open source hardware for building industrial IoT devices easily and quickly. mangOH, which is based on the Qualcomm Technologies 9215 LTE modem, delivers a 90 percent solution prototype out-of-the-box, allowing software developers to easily design customized IoT solutions.
And Samsara is now using Qualcomm Technologies’ MDM9207 IoT LTE modem and QCA9377 chipset in its VG34 IoT gateway. Designed specifically for the IIoT and vehicle tracking industries, the transportation gateway is engineered to deliver faster access than previous generation tech to real-time GPS and sensor data as well as diagnostics, improving efficiency, safety, and customer service.
As the IIoT ecosystem expands to include new sectors and new classes of applications and services, manufacturers must rethink what type of data, and how much of it, should be handled in the cloud. Working with key players in the ecosystem, we’re working to deliver proven low-power, high-performance technologies to meet the varied needs of today’s — and tomorrow’s — IoT gateways.