The Internet of Things: Understanding its Uses

The concept of the Internet of Things emerged over the last decade as computers and wireless devices became so affordable and power efficient that they could be deployed almost anywhere. The Internet of Things (IoT) is not a separate internet but an extension of the existing one. With a low-cost wireless connection, even the most modest computer can own an IP address just like a desktop or smartphone. 

In this three-part series, we’ll provide an overview of IoT concepts, examine various devices and components, and discuss challenges relating to its rapid growth and proliferation.

IoT Adoption

IoT node devices collect sensor data, in some cases including audio and video, and send it to IoT gateways, other devices, or directly to the cloud. Most people encounter IoT in the form of smart home hubs and home automation devices like smart locks, thermometers, and sprinkling systems. Yet, industrial, infrastructure, and government expenditures for IoT devices are already greater than the consumer segment. 

Manufacturers are installing monitoring sensors, utilities are deploying smart meters, farmers are placing sensors on their equipment and in the fields, and governments are starting to invest in smart city IoT devices, such as environmental sensors and traffic control devices. In these scenarios, IoT can help increase efficiency, reduce waste, and avoid downtime with predictive maintenance.

As noted in a recent ZDNet report on IoT, the widespread adoption of IPv6, which expanded the number of IP addresses, was essential to the IoT revolution. The story cites an IDC projection that there will be some 41.6 billion connected IoT devices by 2025. The fact that Network World cites a Priceonomics report that claims there were 50 billion IoT devices already in use in 2020 shows just how fluidly different analysts define IoT.

IoT vs. Embedded vs. Edge Computing

The Internet of Things is often used interchangeably with two other labels that have slightly different connotations. An earlier term still in widespread use is “embedded computing,” which usually refers to industrial automation computers. 

Like IoT gateways, embedded computers typically run a single specialized application. Like IoT gateways, embedded computers increasingly offer wireless connections. Yet, they are more focused on wired interfaces, such as serial, GPIO, USB, or Ethernet, which are used in fieldbus automation controllers. While IoT gateways usually limit themselves to collecting and fusing data from nodes, embedded computers are more likely to also control equipment. Deployments that focus more on monitoring data collection are sometimes referred to as Industrial IoT, or IIoT.

More recently, the term “edge computing” has come into fashion to describe powerful IoT gateways or embedded computers with extensive multimedia or AI support. An edge computer processes more data locally than a typical IoT gateway. They increasingly run AI algorithms for computer vision in so-called “edge AI” systems. 

Boosting local processing helps to reduce communications costs and latency—the lag time between request and response from the cloud that can hamper real-time critical applications. It also ensures the system can continue to run if the internet is disconnected. 

There is also a growing trend toward edge computers that use cloud-native technologies, such as containers, to run more than one app simultaneously. Containers can also improve security and reduce development and maintenance costs when scaling up networks of hundreds or thousands of nearly identical edge computers.

Areas of Overlap

The terms IoT, embedded, and edge computing have considerable overlap and are often used differently by different people. Some embedded devices do not fit neatly in any of the categories, such as digital signage, retail point of sale (PoS), medical computers, robots, drones, and computers that monitor and control assisted driving or autonomous vehicles. 

There are also overlaps with other categories. Smartwatches and wearables can function as sensor endpoints to collect health data, but some can also run apps like a phone. A mini-PC could act as a smart hub, a desktop computer, or a media player. Voice-assisted smart hubs like Amazon Echo also do double duty as streaming audio smart speakers. High-end edge computers, meanwhile, sometimes act like routers or servers.

In the next article, we’ll continue our discussion of the Internet of Things with a look at IoT devices and components.