The IoT and the Road to 20 Billion Devices: How the IoT Coevolves with Its Enabling Technologies

Part 1: A History of the IoT and Its Enabling Technologies

The Internet of Things (IoT) refers to connected devices that merge the physical world with the internet. These connected devices collect data, and then process and translate those data into a service for users’ consumption. In many cases, these devices are invisible, and their only human interface is through communication technologies such as Wi-Fi, Bluetooth, or—in the near future—5G. The cost of processors, sensors, and storage has resulted in an explosion of IoT devices that cover a spectrum of services. This article explores the evolution of the IoT and how it will translate into 20 billion connected devices in 2020.

In 2020, more than 20 billion IoT devices will be connected to the internet, accounting for a market size of about half a trillion dollars. This figure represents quite a leap from the academic novelty of the IoT’s origins. In the 1980s, programmers at Carnegie Mellon University connected a soda machine to their network to be able to remotely identify whether soda was available in the machine and whether that soda were cold. Four decades later, this practical experiment has evolved into a massive market and a huge spectrum of internet-connected product offerings.

And just as IoT devices provide useful services to consumers and companies alike, they drive changes in the technologies that support them. Two key examples include Linux (software) and ARM processors (hardware):

• Linux is a full-featured operating system that maintains an impressive amount of flexibility, running on the most powerful supercomputers and the smallest embedded devices. With the growth of the IoT, Linux has scaled through many distributions to address the IoT market. Examples include Android ThingsOpens a new window ; Ubuntu for the IoTOpens a new window ; and the Yocto ProjectOpens a new window , a Linux distribution that facilitates the construction of a Linux distribution for embedded projects.
• ARM processors are the most widely used microprocessors in embedded and low-level devices such as mobile phones, handheld gaming systems, and the IoT. ARM provides a spectrum of processor types designed to meet the performance needs of target products. ARM processors have enabled products like Raspberry Pi–embedded computers (with Gigabit Ethernet, USB, and even HDMI outputs) and enabled new markets of internet-connected devices (also running the Linux operating system).

The open source Linux operating system and ARM processors have intersected to create an environment in which IoT device development is trivial and requires minimal time to market. An example of the innovation that Linux and ARM make possible is the MagicMirror²Opens a new window project. This project is a smart mirror made up of a double-sided mirror in front of an inexpensive monitor powered by a Raspberry Pi board running Linux. The monitor shines through the mirror, providing notifications and other information. The project has created a massive DIY effort for smart mirror development, and a large community of developers now exists for modules that add new capabilities to the project, such as facial or gesture recognition.

Similarly, the cloud is an enabling force for IoT development. Rather than purchase servers, storage and networking equipment, and private internet access, public clouds enable consumers to purchase these resources and scale them dynamically based on their need (with low up-front costs but higher costs in the long term). The result is a unique opportunity for IoT developers, who can focus their resources and attention on IoT device development rather than ecosystem build-up and management—important when the focus is on time to market rather than end-to-end infrastructure development and integration.

Part 2: IoT Devices and Markets

Anatomy of an IoT Ecosystem

The IoT represents a diverse set of products and architectures, but one common theme of an IoT ecosystem has appeared. It consists of an IoT device that works with a remote cloud mediated by a gateway device:

• The device represents the “thing” in the Internet of Things. it provides a service to a user or entity and may or may not provide a direct user interface (UI), but it typically communicates over a wireless protocol (such as Wi-Fi or Bluetooth). The device is the core of the IoT ecosystem and the focus of the value the IoT delivers.
• The gateway is the conduit for one or more IoT devices in a given area, such as a home or warehouse. The gateway mediates communication between the IoT device and the cloud; it may also represent an aggregation point through which multiple IoT devices are managed. The gateway can also serve as the end point for IoT device data and management if no cloud is required.
• The cloud is the storage and computation infrastructure for the IoT ecosystem. The cloud automates and orchestrates the management and monitoring of IoT devices as well as the analytics of the IoT device data. The data analytics can be used for many purposes, such as billing or optimization, and—depending on the device—may be exposed back to the IoT end user. The cloud is an ideal consumer of IoT data because it more easily scales to IoT deployments (both expanding compute and storage resources as IoT devices proliferate and shrinking based on the dynamic demands of those devices).

The introduction of 5G alters this ecosystem by limiting the need for gateways when devices are 5G-enabled. Not all IoT devices will be 5G capable, however, so the ecosystem described here will remain a common architectural pattern, with gateways benefiting from 5G communication.

IoT Product Characteristics

IoT products span the spectrum of embedded devices. But, some have key characteristics that differentiate them from typical devices:

• Communication. The device has either direct network connectivity through Wi-Fi or other wireless or wired communication to a network-capable device, such as a gateway.
• Sensors. Using one or more sensors, the device collects data. Sensors may also support a UI, such as voice recognition, gesture identification, or facial recognition.
• Data. The device shares the data that sensors have collected and fused to the cloud for immediate or deferred processing.
• Analysis. Either locally or in the cloud, data are analyzed for some application-specific purpose. Data can be analyzed in isolation for a given instance or in the context of a population of data to identify insights not possible with a single data set.
• Actuation. To effect changes at the IoT device, the IoT system may include actuators that are managed locally or remotely based on the data analysis.

The key differentiator for IoT products is their connectivity to the internet. That connectivity provides the means to collect data for analysis at the gateway or cloud or provide some effect at the IoT device. That process is a key goal for IoT devices.

IoT Markets

IoT markets vary and cover the range from connected consumer electronics to manufacturing and transportation. Although not exhaustive, the following market segments define some of the most important or fastest-growing segments in the IoT world:

• Building and home automation. The benefits of the IoT to building and home automation are clear, and this segment represents one of the fastest-growing within the IoT world. An initial driver for this segment was energy efficiency, but it has now extended to security systems, entry systems, connected entertainment, light control, video systems, doorbells, and a variety of other applications. Integration of these solutions into an intimately interconnected system will represent a key future challenge.
• Smart energy. Covering smart grids and the goal of increasing the efficiency and sustainability of large-scale power management, this future market is key within the IoT. Transforming energy management in the context of delivery and improving use by consumers is important for reducing overall carbon dioxide emissions. Smart energy is closely related to smart cities, where the IoT is used to gather data for the efficient management of resources and services.
• Smart retail. This category describes a range of solutions based on technologies for optimizing commerce and improving the consumer shopping experience. It includes radio-frequency identification (RFID) tags and equipment for tracking, beacons that permit pushing notifications to consumers’ smartphones while shopping, and changing the point of sale into a decentralized experience for customers (such as being able to check out anywhere).
• Connected health. This category refers to the consolidation of patient health monitoring and management, both in hospitals and at home. It covers management and monitoring of patients, medications, care, equipment, and overall optimization to improve care and reduce costs. With medical errors representing the third-leading cause of death in America, connected health could help reduce or eliminate these preventable mistakes.
• Automotive. This category includes the various technologies that improve our driving experience, making it safer and in aggregate freeing us to do other things. From systems that use facial recognition to detect a driver’s emotional state (distracted, sleepy, or even angry) to reducing a bumpy ride by using sensors and hydraulics to counteract contours in the road, automotive IoT is an open field. As our lives become even more distracted, self-driving vehicles and associated IoT sensors and actuators will result in a rapidly growing market.
• Transportation and logistics. This category covers the various aspects of supply chain management, tracking vehicles and their cargo in real time; monitoring deliveries; and optimizing routes to improve efficiency, increase safety, and reduce overall cost.
• Platform. Platform refers to applications and analytics that provide management over IoT devices along with data analytics to measure and optimize such management at scale. This market includes end-device management, enterprise-class software for management and monitoring, analytics platforms to support real-time decision making, and the scalable cloud infrastructures that provide the basis for their operation.

Innovation continues in the IoT, and new markets will appear as new applications for this ecosystem are discovered.

IoT Building Blocks

As wide and varied as IoT markets are, common IoT building blocks are helping drive the IoT toward 20 billion devices. These building blocks consist of technologies that provide communication, processing, and sensing.

For communication, more than half of the IoT devices in use today connect to one of three technologies:

• Bluetooth. Bluetooth operates in the 2.5-gigahertz spectrum over short distances and requires little power (which is improved even further with the Bluetooth low energy standard).
• Z-Wave. Z-Wave operates at longer distances (up to 90 meters open air) and is commonly used in home automation applications. It implements mesh networking (also called wireless ad hoc networks) and is fully standardized for interoperability between manufacturers.
• Zigbee. Zigbee implements mesh networking but is cheaper than other wireless technologies. It supports short-distance, low-power communication at low bandwidths and is common in home automation and medical electronics.

5G is also on its way, with features focused on the IoT. It will be a game changer for the IoT, but it’s too early to tell when this the technology will be adopted widely.

A rich set of single-board computers addresses the spectrum of IoT product needs. Raspberry Pi is one key example that supports deeply embedded devices; it also meets the requirements for IoT gateways to aggregate and communicate in the cloud. With its support for Linux (called Raspbian Linux and derived from the Debian Linux distribution), Raspberry Pi addresses any current application need.

Sensors merge an IoT device with the real world to provide a way to sense and interact. Sensors vary and, depending on the application, can measure (among other things) pressure, temperature, images, audio, motion, gas, and chemicals.

In the context of logistics, RFID sensors focus on identification of assets. RFID is also a communication technology, but it’s much more specific to tracking applications. Passive RFID tags have no power; instead, they rely on external interrogating radio waves for their operation. Active tags can use a local power source and operate over longer distances. RFID tags contain data that can be transmitted to aid in identification of an asset for tracking and management.

Part 3: Summary, Outlook, and Resources

Hardware, software, sensors, and applications have intersected to support the massive influx of IoT development that is actively creating new products and markets. The Linux operating system enables fast time to market by enabling IoT companies to focus on their value rather than the development of product operating system platforms. With public (and private) cloud infrastructures, the ability to manage massive numbers of IoT devices is now in the hands of anyone with the desire to create.

The Challenges of the IoT

The challenges the IoT faces are beyond development and deployment—namely, device security and data privacy. A quick look at the Shodan IoT search engine reveals unsecured IoT devices such as cameras, illustrating this problem clearly. Device security is difficult enough, but when a device that can be secured is left unsecured, then it’s easily exploited. The larger issue with IoT security is that a product that is exploited can be multiplied into a botnet and weaponized for distributed denial of service attacks.

Data privacy is the other key concern if the IoT proliferates as predicted. When audio or video information, for example, is stored from a doorbell or security camera, this information must be protected and used in a way that is well documented for consumers. The European Union (EU) recently passed a regulation on data protection and privacy called the General Data Protection Regulation (GDPR), which ensures both data protection and the right to be forgotten (data erasure). The GDPR also grants users the ability to access their stored information to prevent misuse of data.

A final challenge for the IoT is the ability of IoT devices to interoperate. Many companies build silos so that their products work in isolation with their specific products, but the future of the IoT relies on interoperability and the means to deploy devices between systems using standards that define mechanisms for both communication of data and the format of those data. Device controls could also benefit from a standard management interface that is rich and scalable across application domains.

Resources

 Category	 Resource	 Description
 Linux	 Anatomy of the Linux KernelOpens a new window	 “Anatomy of the Linux Kernel” provides a short history and architectural decomposition of the GNU/Linux operating system.
Raspberry Pi	 Raspberry PiOpens a new window	 The Raspberry Pi Foundation’s mission is to put the power of computing into the hands of people everywhere.
IoT Gateway	 What Is an IoT Gateway?Opens a new window	 This post simply describes the idea of an IoT gateway and provides various use cases.
Cloud	 Cloud computingOpens a new window	 Wikipedia’s article on cloud computing covers the models of cloud computing and storage with their architectures and challenges.
Raspbian Linux	 Raspbian Linux websiteOpens a new window	 This home of the Raspbian Linux distribution includes packages and installation instructions for Raspberry Pi.
Bluetooth	 Bluetooth websiteOpens a new window	 Find specifications, resources, and articles on developing with Bluetooth technology.
Z-Wave	 Z-Wave websiteOpens a new window	 The Z-wave website focuses on home automation and offers a useful blog that describes Z-Wave applications.
Zigbee	 Zigbee AllianceOpens a new window	 The Zigbee Alliance’s website includes developer resources and solutions related to Zigbee.
Shodan	 Shodan search engineOpens a new window	 Shodan is the world’s first search engine for IoT devices and demonstrates the issues with IoT device security.
GDPR	 General Data Protection Regulation websiteOpens a new window	 The GDPR site contains rules for the protection of personal data inside and outside of the EU. Although this law is enforced within the EU, it covers international organizations that offer services to users within the EU.
RFID	 Wikipedia’s RFID entryOpens a new window	 Wikipedia’s article on RFID covers the technical details of the technology, the standards that govern it, and extensive resources.