The Internet of Things Hits Its Stride

By John Donovan, Mouser Electronics

The current Internet of connected humans through social media is quickly and quietly morphing into an Internet of interconnected objects. In October 2013, IDC predicted that the Internet of interconnected objects will be an $8.9 trillion dollar market in 2020 with 30.1 billion autonomous devices connected. With the number of connected machines projected to outnumber the earth's population by 4:1 by 2020, the future Internet will be primarily the Internet of Things—or, with due deference to humans, the Internet of Everything.

The Internet of Things (IoT) is enabled by machine-to-machine communication between devices, which in this case are communicating over the Internet. The connected devices have embedded intelligence that enables them to monitor and analyze data and take action without human intervention, although humans set up, monitor, and ultimately control these networks. Machine intelligence makes IoT systems a lot more efficient by performing operations that can be done more quickly, consistently, and reliably than systems that require humans to sample, analyze, and act on information.

Everything old is new again

Like all seemingly new ideas, the Internet of Things has some history. As far back as 1999, Bill Joy (inventor of Berkeley Unix and Sun Microsystems) described a "device to device (D2D) web... an Internet of sensors deployed in mesh networks" that would embed machine intelligence in everyday life. The idea lay dormant for ten years until MIT's Kevin Ashton coined the term Internet of Things to describe how digitally tagged objects could be networked together, interacting without the need for human intervention. In Ashton's words, "If we had computers that knew everything there was to know about things— using data they gathered without any help from us— we would be able to track and count everything, and greatly reduce waste, loss and cost."

Working out of MIT's Auto-ID Center, Ashton's initial concept was to tag everything and possibly everyone with RFID tags or possibly bar codes. Enabling "tracking and counting" is fine for tracking inventory, but it is not for control or communication that requires embedded intelligence and connectivity— both part of Joy's vision of a D2D web that enables "ubiquitous computing".

Industry started putting the infrastructure in place for the Internet of Things with the introduction of supervisory control and data acquisition (SCADA) systems, such as remote terminals and programmable logic controllers, as early as the 1970s. A central computer would interact with sensors and actuators embedded in machines along an assembly line to control the operation of the line. Distributed lines led to distributed systems and later networked SCADA systems with more and more intelligence being built into peripheral devices.

With the introduction of the Internet, the result has been devices that are able to sense, compute, and communicate—an Internet of distributed objects that harvest information from the environment and independently interact with it. Moving beyond the factory floor, Internet of Things devices—each with embedded sensors, actuators, communication, and computing elements—can create a smart environment with a wide range of applications in healthcare, public safety, transportation, utilities, and the home.

There's an app for that

Utility companies were early participants in the Internet of Things in the form of the smart grid. The smart grid is essentially a large, high-voltage communications network, although the backbone network might be cellular, Wi-Fi, or satellite based. Distributed systems throughout the grid continuously monitor usage and reroute power or respond to problems as needed. At the far end, smart meters continuously monitor consumption, enabling both utilities and end users to better manage their resources.

Smart devices in the home respond to remote directives to reduce power during peak loads or to power up later when demand on the grid slacks off. Smart air conditioners, washing machines, and other appliances enable better home energy management and allow electric companies to avoid blackout and brownouts without additional investment in equipment.

Health and fitness buffs already wear monitors that record their heart rate and the distance they run, coupling that to a PC to analyze the results. Wearable wireless medical devices include accelerometers to warn of falls, EKGs for heart monitoring, and insulin pumps and glucose monitors for diabetics. Each of these devices can connect to a mobile phone or PC via Bluetooth and an Internet of Things back end to upload data to servers at medical offices. End-to-end solutions in health monitoring are quickly becoming a reality with standards and regulatory bodies no longer dragging their feet. Early in 2014 the Continua Health Alliance will make public its lightweight protocol that works with mobile devices and GSMA has already made mobile health one of its target growth areas.

Get smart

At the heart of most IoT implementations are low-power wireless sensor networks (WSN) that connect to the Internet. A typical WSN node contains one or more sensors, A/D and possibly D/A converters, an MCU, an RF transceiver, and a power supply (often a battery). While a node may contain enough intelligence to respond to environmental changes, nodes usually send data to a distributed or centralized system for analysis. Active RFID, with its very limited processing and storage capabilities, comes in at the low end of the WSN range. More intelligent WSNs can perform high-speed data collection, processing, and communications, moving a great deal of control out to the node level.

As cellular and wireless connections increase, cloud-based computing is becoming central to the Internet of Things. The cloud provides scalable storage for the massive quantities of data provided by WSNs as well as the analytical tools required by human users to make sense of all that data. The cloud can provide the necessary computing resources and services to manage large IoT networks. Cloud-based software-as-a service (SaaS) is key to scaling IoT applications.

However, scaling SaaS itself is a problem due to a lack of standardized cloud APIs. Hardware platforms from Arduino and Microchip along with Google's accessory development kit provides developers with an application platform-as-a-service (PaaS), with a runtime environment and a set of APIs that enable developers to build custom applications using multiple programming models.

Intel supports the idea that a common hardware/software platform with industry standard APIs and a large supporting ecosystem are key to making the Internet of Things work smoothly. Intel is promoting the advantages of IoT gateway solutions based in its low-power E3800 Atom™ Processor series as well as the new Quark SoC in the new open source Galileo board.

Freescale® is also a strong believer in the Internet of Things. Freescale's ARM® Cortex^TM-M0 based Kinetis KL02 MCU offers a combination of speed, storage, and energy efficiency that makes it a logical choice for low-power wireless sensor nodes.

Challenges and issues

One obstacle to connecting billions of devices to the Internet is the shortage of remaining IP addresses using IPv4. Also, IPv4 can only identify a group of co-located sensors by location, but not individually, creating a roadblock between the gateway and the wireless sensor devices. IPv6 partially answers the scaling and mobility problems, but it will not be adopted overnight. Even then, WSNs run a different software stack that cannot be addressed directly from the Internet. The gateway needs a Uniform Resource Name (URN) subnet to address individual sensor nodes, each of which has a uniform resource identifier (URI). Sensor nodes will be accessible by URL but only addressable by URN to the URI. Clearly, implementing IoT networks is no simple matter; hence the appeal of PaaS.

As the Internet of Things expands it raises safety and privacy issues. IoT gives a virtual presence to any number of physical objects. It has been demonstrated that current SCADA systems controlling the electrical grid can be hacked; The Stuxnet worm brought that issue to public consciousness in 2010. The smart grid and all its diverse, interconnected devices represent an extremely large attack surface. The safety issue is being addressed aggressively at the system, software, and hardware levels, but it is a problem that can only be mitigated as it will not go away.

Pervasive computing also raises privacy issues. Having security cameras monitoring a crowded subway station is one thing, but having them monitor a person's every move is quite another. The same GPS/cellular technology that enables a person to navigate to another city or find a nearby Starbucks using their cell phone also constantly tracks their location. Geopositioning is useful for fleet management and tracking containers, but tracking the day to day movement of people raises privacy issues that, like the security ones, are not easily dismissed.

Back to the future

While enabling the delivery of far more efficient services, the Internet of Things is changing the nature of the Internet itself. Web 1.0 consisted of static web pages. Web 2.0 ushered in dynamic web pages and interactive social media. Now the simultaneous explosion of connected devices and cloud-based software services is ushering in Web 3.0, where there will be the ability to interact seamlessly with machines—so much so that the current generation of children will take it for granted. Bill Joy's D2D web will finally become a reality and be fully integrated into the Internet over which humans will continue to interact.

The future Internet will be fully integrated with the Internet of Things. Are you ready for it? Your machines are.

John Donovan is editor/publisher of www.low-powerdesign.com and ex-Editor-in-Chief of Portable Design, Managing Editor of EDN Asia, and Asian editor of Circuits Assembly and Printed Circuit Fabrication. He has 30 years experience as a technical writer, editor and semiconductor PR flack, having survived earlier careers as a C programmer and microwave technician. John has published two books, dozens of manuals and hundreds of articles. He is a member the Association for Computing Machinery (ACM) and a Senior Member of the IEEE. His favorite pastimes include ham radio, playing with his kids and scouting Texas for the best BBQ joints.