A Successful “Internet of Things” Hinges On M2M

Network everything! That’s the battle cry of the movement called the Internet of Things (IoT), or according to some, the Internet of Everything (IoE). Whatever the name, a major campaign is now underway to push along its long-awaited implementation. Most visible within this undertaking is machine-to-machine (M2M) connectivity.

M2M is IoT. It’s taking time to become accustomed to this idea, but progress is being made in achieving the ultimate goal—a network connecting practically all physical devices. Some call the IoT movement the fourth generation of the Internet.

The first generation back in the 1970s and 1980s involved the ARPANET and the linking of the military, government agencies, and selected universities throughout the U.S. The second generation in the 1990s ushered in the AOL phase with its extensive e-mail and beginning Web browsing. Social media dominates today’s third generation.

IoT represents the fourth generation, dubbed by many as the “networking of everything else.” It’s now technically and economically possible to put virtually everything on the Internet, setting up what looks to be a very interesting future with all of the potential benefits, downsides, and unintended consequences. 

Network What?

The whole reason for the IoT is to be able to monitor and/or control every device of interest or importance. That means telemetry and automation on a mass scale. In a pure M2M approach, one machine or thing talks to another machine or thing independently of human interaction. Other IoT scenarios, though, consist of M2M plus machine-to-person or person-to-machine connections. Some of these connections will prove highly valuable, while others border on the ridiculous.

To illustrate, connecting your toaster to the Internet appears silly on first glance. However, it may not be such a bad idea after all. For example, you could check that toaster to check on an aging parent’s diet. You could remotely access the toaster, which would tell you the date and time it was last used. If the toaster is used daily, you could assume the parent is eating breakfast, at the least, and is still active.

Excessive toaster use may indicate visitors or overeating. Other information to glean can include the condition of the toaster, e.g., is it heating properly and is the temperature setting high, low, or medium? Or perhaps it could indicate that the toaster is defective. Toaster inactivity may also indicate a problem with the parent. The bottom line is that there’s more to these connections than initially appears obvious. Context is important.

Nestle’s Nespresso SA’s M2M-connected professional coffee machine also initially appears to be a silly idea, yet it turns out to be a good business decision that leads to both customer and company benefits (Fig. 1). The M2M service makes sure the machine is well maintained. It can monitor daily use, correct temperature setting, vibration, and pressure, and thereby ensure good quality output.

Also, the company can be informed about the need to order coffee supplies. When the need arises, the service does it automatically. On top of that, the process detects problems and/or anticipates them from collected usage data. Perhaps it’s not meant for every home Mr. Coffee, but on an expensive professional coffee maker, it ensures customer satisfaction and continued repeat business.

The Value Proposition

The IoT automatically collects data from some connected asset and guarantees its delivery to a person or computer. Then beneficial decisions can be made from that data. However, the value goes beyond the data or facts. Knowledge gleaned from the data is quite valuable if it can be applied.

As a result, companies must identify the problems or challenges—and their value. It could be a creative solution to a long-ignored problem or a way to provide a new revenue stream. Therefore, companies have to be aware that M2M is a viable option and possible solution to many problems. Costs are now low enough to allow for M2M implementation without breaking the bank and providing a real return on investment (ROI).

What Makes IoT Possible?

The availability of low-cost sensors and communications modules has made it economical to embed Internet connectivity into many devices. Couple that with the new IPv6 protocol, with its 2¹²⁸= 3.4028236692093846346337460743177e+38 addresses, and there’s sufficient identification for virtually all objects worldwide. Not all connected things will need their own Internet Protocol (IP) address. Nonetheless, we’re covered for decades to come.

In addition, M2M services simplify implementation and deployment of IoT apps. M2M service companies like RACO Wireless help customers build their M2M networks, provide connectivity services, and collect, store, and analyze the acquired data.

Furthermore, wireless networks continue to expand and accelerate without compromising reliability when communicating data to and from things. Also, more development platforms are available, helping potential users create their own applications software.

Diverse Applications

The term “diverse” seriously understates the variety of current and potential applications. Some have been around for a long time. GM’s OnStar vehicle service, for instance, exemplifies the key attributes of an M2M process. It can determine if you had an accident, locate you with its GPS capability, and automatically call 911. And its ability to unlock car doors (quickly mitigating the maddening situation of keys locked inside the car) has proven to be quite handy.

Another example comes from Amazon’s Kindle E-reader. Its built-in cellular service provides a way for you to shop for a book or other item, then download it in a few seconds. It’s fast, convenient, and definitely addictive to some.

Home-security services like ADT already offer a wide range of security options. They use the Internet and your cell phone to monitor home breaches, fire, water and flooding, and light and thermostat control, as well as make video observations.

Most IoT researchers divide the applications into categories:

• Asset tracking and inventory: Owners of valuable pieces of equipment, such as bulldozers, front-end loaders, road graders, and tractors, can stay on top of their location and condition. The presence of valuable inventory also can be determined, whether it’s large objects on shipping pallets or small items on shelves.
• Shipping and location: M2M, with the use of GPS, enables asset tracking and can provide an update on its general condition. Commercial trucking companies have used M2M for years to keep track of their trucks and their valuable cargo. It answers questions such as “Will the delivery be on time?” “Is the refrigerated van at the right temperature?” “Is the truck broken down?” “Is the driver resting?” or “Is it time for an oil change?” It also provides information on the location of the shipping container that was recently offloaded a ship.
• Vending-machine service: M2M allows vending-machine owners to determine the condition of the machine and whether it’s out of product. That can save truck rolls and unnecessary preventative maintenance visits, resulting in an optimized asset.
• Home: Home area networks (HANs) may be connected to the Internet via their high-speed cable TV or DSL line, or possibly via a cellular connection. If a local utility monitors and/or controls that HAN, it becomes an M2M node. Smart home monitoring for energy savings is the main application.
• Vehicle: Vehicle M2M already exists as a customer safety and convenience connection, such as the GM OnStar. Many other auto companies offer similar services. If the car or truck is an M2M node, it can provide all types of data about vehicle status from a multitude of sensors, many already on board, and report it to the driver. This data may be collected in a “black box” for use in dealer diagnostic, maintenance, and repair, or in case of an accident. Many possibilities abound, including the ability for nearby vehicles to communicate independently with one another for safety purposes. Communications with roadside information services is another possibility.
• The Smart Grid: The Smart Grid is certain to expand its use of M2M. It uses intelligent computers and communications to better monitor and control the electrical grid. Sensors and embedded controllers on high-voltage distribution towers, substations, and even individual neighborhood transformers can better detect and manage the grid for energy savings, as well as report outages. In many instances, the grid will be able to automatically adapt as needed without human interaction.
• Farming/ranching: M2M sensors are able to monitor crop conditions, such as soil moisture and surrounding temperature. Automatic watering systems can be monitored and controlled, too. In terms of ranching, radio-frequency identification (RFID) tags applied to animals aids in counting and, in some cases, location.
• Medical and fitness: Major potential applications surround the world of healthcare. Patient monitoring will go beyond the hospital and into the home, or anywhere a communications link can be established. Doctors can monitor conditions to diagnose problems and practice preventative medicine after receiving early warnings of potentially lethal events. Sports and fitness may benefit by collecting health and performance data of athletes or those who run or otherwise exercise. Rehabilitation is another possible application source.
• Environmental monitoring: It’s currently in place, but M2M advances will cut costs and simplify the process. Weather conditions are already fully monitored. Now other conditions like air or water quality will receive closer scrutiny.
• Industrial automation: This also is already occurring, at various levels, in factories and process control plants. Manufacturing or processing operations are currently monitored and controlled internally, but Internet connectivity adds the potential for remote monitoring, reporting, and control.
• Security: Today’s security systems use myriad sensors and video cameras for internal detection and reporting of security breaches. Adding an Internet connection greatly enriches their potency. A number of security companies, in fact, already implement such systems, enabling clients to monitor their home or business on a cell phone.
• The M2M Vision
• The entire concept behind M2M involves large numbers of devices delivering information and collectively making decisions without human interaction. Add the human element to machine interaction, and it markedly broadens the scope of M2M possibilities. The ultimate goal is real-time monitoring and control, which implies each M2M node embeds intelligence that interacts with the intelligence at remote collection points and related severs (Fig. 2).

• Specifically, the end M2M nodes contain one or more sensors that report physical conditions to a remote site or are used with local embedded intelligence. Some nodes will also have an actuator to initiate action based on sensed conditions, or in response to some remote command.
• An example of a node is a street light that reports its status. For instance, it can simply report that it’s working or burned out. Furthermore, a light sensor could detect a time to turn on or off, or the turn on and off times may be provided remotely. (The actuator turns the light off and on at desired times.) The node also includes a communications transmitter and receiver. Substantial energy and maintenance cost savings have made these nodes popular within city lighting systems.
• Parking meters represent another example. A city can monitor usage, time of day activity, funds collection, and violations. As a result, it’s possible to optimize meter and parking-space usage.
• Most nodes will probably report to a nearby aggregation point that serves multiple sensors. The communications link may be Wi-Fi, but other wireless options such as cellular are possible. Think of cellular as wireless backhaul for M2M. This aggregation point then connects to a gateway setup for the application. The gateway may handle multiple aggregation points. In some applications, though, the aggregation points may not be necessary. Again, the link can be Wi-Fi or another wireless connection.
• Another real possibility is that the application may have multiple personal area networks (PANs) from ZigBee, Wi-Fi, or Bluetooth wireless sensor network connections. These might report to a separate gateway. The gateway then reports up to the next level via a cellular connection to the Internet cloud. An Internet link connects to a server that hosts the middleware implementing the application. Other links are possible, too, depending on the scope of the application.

Internet Connection Technologies

Practically every communications technology out there could implement IoT and M2M. Today, cellular connectivity dominates the M2M space, but all other wireless and wired technologies are involved as well.

With the current array of semiconductor devices at one’s disposal, a complete data (no voice) cell phone can be packaged small enough to embed into almost anything. Cell-phone module makers offer a wide range of data radios using GSM/GPRS/EDGE or cdma2000 1xRTT for low-speed applications (e.g., sensor data collection). Though most of today’s M2M applications operate at very low data rates, faster modules using WCDMA, HSPA, or even Long-Term Evolution (LTE) have arrived to handle video and other high-speed applications.

Wi-Fi 802.11 connectivity also will continue to play a role. Its high speed and low cost make it attractive for some applications. Moreover, it features a longer range and doesn’t load the cellular network, at least directly. It requires one or more accessible hotspots, which are easy to build. Even existing access points can serve an application. (Some of the nodes will have GPS location capability.)

Then there’s ZigBee and other short-range wireless devices that can be formed into wireless sensor networks. Z-Wave and a mix of industrial, scientific, and medical (ISM) band modules also will become parts of these networks. In addition, Bluetooth can form personal-area networks (PANs) and thus fit certain applications. For instance, Bluetooth Low Energy 4.0 and similar ultra-low-power versions suit medical and fitness applications particularly well.

ZigBee and other 802.15.4 standard wireless devices are ideal options thanks to their low power consumption and ability to form wireless mesh sensor networks. The 6LoWPAN standard from the Internet Engineering Task Force (IETF) permits IPv6 packets to be carried over 802.15.4 low-power devices to connect billions and billions of devices.

Two wireless technologies sure to flood the IoT space are RFID and near-field communications (NFC). RFID can replace standard bar coding ID in many applications. RFID tags are cheap and don’t use local power.

Passive RFID tags get their power from a nearby reader that transmits a signal used by the tag to power up its transmitter. The transmitter then sends an ID code back to the reader. Read distances are very short, from a few feet to a few inches. Range depends on the radio technology—tags operating in the 125-kHz, 13.56-MHz, and 900-MHz ranges are available. Active tags use small batteries to extend the tracking range on more expensive items.

NFC, like RFID, uses the wireless near field (as opposed to the far field) for communications. Range is limited to inches, but that’s still useful for many applications. An international standard at 13.56 MHz promises to be the next big cell-phone addition, enabling the implementation of the “digital wallet.” NFC-connected smart phones can replace credit cards for making payments—just tap the vendor’s NFC reader with your smart phone to conclude a sale. The NFC component may also be used for building access or transportation.

Some smart phones already integrate NFC, but it’s still at the early stages. As soon as retailers, credit-card companies, and other vendors design their systems and install readers, look for their explosion across the IoT.

Still Wired In

Though wireless connections to the Internet will dominate IoT and M2M applications because of their convenience, wired technology offers plenty of good options for some applications. For example, power-line communications (PLC) links are fast and easy if there’s an available ac power line.

Other possibilities include HomePlug and G.hn for high-speed applications, and G3, PRIME, et al, for the slower data needs. All use orthogonal frequency-division modulation (OFDM), so they hold up well against the noise and high attenuation of the ac mains. For a more detailed look at all of these options, see “Power-Line Communications Emerge As A Core Networking Technology” and “New Products Jump On The PLC Bandwagon."

Wired Ethernet is another option. With Ethernet local-area networks (LANs) virtually everywhere, it’s possible that such a network will ideally serve the communications needs of some M2M applications.

M2M: A Work In Progress

M2M and IoT are in their infancy, but recent market data shows they’re maturing rapidly (see “No Shortage Of M2M Products”). ABI Research reports sales of 33.9 million M2M modules in 2010, which jumped to 42.5 million in 2011. Projected 2012 sales are expected to come in around 53.9 million modules, while longer-range estimates show 101.5 million modules sold in 2015. That equates to a 25% annual growth rate.

Francis Sideco, senior principal analyst at market research firm IHS, projects that cellular M2M modules will catapult from 20 million units in 2010 to 250 million in 2015, along with a 41% compound annual growth rate (CAGR) increase during that same period. He says that while growth is good now, it will really blossom into the desired hockey-stick curve once some standardization helps overcome the current fragmentation. Check out IHS’s latest report M2M 2012: Fragmented Requirements Still Biggest Challenge for Internet of Things.

In other research, Beecham Research’s latest report projects a market of $2.2 billion by 2014 for M2M enablement services. Robin Duke-Woolley, CEO of Beecham, sees growth opportunity for these layers above the network layer, enabling the multitude of diverse applications.

Pyramid Research’s new report for mobile network operators (MNOs) like AT&T shows M2M airtime revenue should grow from $2.2 billion in 2010 to $7 billion in 2016. MNO success in M2M hinges on flexible pricing, according to Pyramid. Clearly, MNOs are looking for new revenue streams due to peaking voice-call revenue and the near saturation of smart phones.

A new report from Juniper Research says that smart metering will account for over 40% of the 400 million M2M-connected devices by 2016. It expects M2M-service annual revenue to eclipse $35 billion by 2016. Juniper further predicts that enterprise M2M connections will increase from 81.8 million in 2011 to 217.3 million in 2015.

But like any emerging technology movement, glitches pop up that can curtail the growth pattern. Major among them is the lack of standards. All wired and wireless standards are covered, but most M2M players agree that the greatest need is some standard protocol to aid in interoperability and more widespread acceptance. The current fragmented situation discourages and slows investment and adoption.

Help is on the way, though. Most global standards development organizations (SDOs) recognize the need for some kind of universal M2M service standards. Organizations such as ARIB, ATIS, CCSA, ETSI, TIA, TTA, and TTC are undertaking an initiative to build an M2M service layer that’s embeddable in hardware and software. There’s no set timeframe as of yet, but work is underway toward global harmonization and consolidation.

The European Telecommunications Standards Institute (ETSI) recently published an M2M service standard for platforms to manage the complexity of multiple M2M technologies and services. It’s designed to provide a complete horizontal service layer for M2M applications.

The TR-50 Smart Device Communications Architecture standard from the Telecommunications Industry Association addresses common requirements and interoperability of M2M devices. The specification works across both wired and wireless transport layers, as well as covert IP-enabled applications covering different types of M2M vertical markets.

One interesting development worth watching is the Message Queuing Telemetry Transport (MQTT) asynchronous inter-device communications protocol. The open royalty-free protocol, invented in 1999 by Andy Stanford-Clark of IBM and Arlen Nipper of Eurotech, targets limited low-bandwidth, high-latency devices typical of IoT/M2M. Check out www.mqtt.org for details.

Wireless spectrum availability presents another sticky problem. With billions of new wireless nodes coming online, is there enough spectrum to handle the traffic? As the cellular operators roll out their 4G/LTE networks, plenty of capacity will be available. But how long will that last?

Increasing demand for more mobile video will quickly eat up all that new bandwidth. While M2M services generally don’t require high-speed capability, the fact that carriers will be dealing with millions of new nodes will certainly make an impact. Simply put, further M2M expansion will require more spectrum. The Federal Communications Commission (FCC) has promised more spectrum, but we will have to wait and see.

In the meantime, many feel the white-space spectrum may be ideal for M2M. White spaces are those unused 6-MHz TV channels between existing channels. The available channels vary from location to location, depending what local TV channels are occupied. This unlicensed spectrum is available across the country and fits well with the low-speed, low duty-cycle operation of M2M nodes. Expect more M2M-related white-space products to arrive soon.

Finally, keep in mind that security issues will enter the fray. Most of today’s wireless systems have security provisions that should suffice for all but the most sensitive applications. When the need arises, security will be handled at the application level.

No Shortage Of M2M Products

Hundreds of chips, modules, and other devices target M2M and the IoT movement. Below are several representatives of this group.

Sierra Wireless makes a mix of cellular modules for embedding as well as gateways and software. For example, the SL6087 is a 2G GSM/EDGE module for low-speed uses (Fig. 1).  Other modules for faster 3G technologies integrate HSPA+ and cdma2000 EV-DO Rev A. LTE modules are now available, too.

Most of these units have internal GPS receivers and related software support. They operate from 3.6 V and feature a variety of interfaces including UART, USB, SPI, I²C, GPIO, and even ADC and PWM. An ARM9 processor is included with these modules, which use AT commands for control.

Sierra Wireless’ GX440 gateway targets high-speed M2M applications (e.g., video surveillance) and operates with LTE 4G networks (Fig. 2).  Its prime mode is LTE on 700 MHz, but it will also fall back to EV-DO Rev. A, EV-DO, or cdma 1xRTT. The gateway, which has built-in GPS, runs an ARM11 processor. Applications interfaces include TCP/IP, UDP, and HTTP. Among the host interfaces are 10/100 Base-T Ethernet with an RJ-45 connection, RS-232, and USB. An onboard IPsec SSL VPN client provides security.

Development software becomes an important part of building any M2M application. In some cases you’re on your own, but Sierra’s ALEOS Application Framework software includes an integrated development environment (IDE), subroutine libraries, and a comprehensive set of tools for building a custom M2M application. ALEOS, which will initially run on the AirLink GX400 and GX440 gateways, makes it possible to tie together the embedded modules to the gateway for any application.

The HE910 embeddable wireless module from Telit Wireless Solutions is 3G HSPA+ compatible and covers all five cellular bands (850/900/1700/1900/2100 MHz), making it usable worldwide (Fig. 3). The device also supports 2G GSM/GPRS/EDGE.

To implement Wi-Fi M2M, B&B Electronics developed the Airborne 802.11b/g Access Point (APCG-Q5420) complete with an Ethernet port and two serial ports (Fig. 4). An embeddable module (APMG-Q551) is shown in the lower right. B&B’s devices target industrial applications.

ISM-band wireless also can implement M2M. Microsemi’s ZL70250 is an example of a sub-1-GHz ISM-band transceiver with extremely low power consumption, suiting it for continuous monitoring of sensors (Fig. 5). Thanks to its Z-Star protocol stack, the device can support a star network with CSMA in wireless sensor networks. It’s easily powered by a coin cell or energy-harvesting sources.

For years, Lantronix has been embedding wired Ethernet connectivity into products. Its xPico embedded device server enables Ethernet connectivity on virtually any device with a serial interface (Fig. 6). The module connects to the application’s processor via a generic asynchronous serial interface with CMOS levels of 3.3 V and data rates up to 921.6 kbits/s. GPIO also is available.

On the network side, the interface is standard 10Base-T or 100Base-TX with auto-sensing. It offers full device server software with a full IP stack. There’s no need to write any code to use it. Encryption is 256 AES, and there’s a built-in Web server. Lantronix offers other options, such as the PremierWave EN device server with Wi-Fi 802.11a/b/g/n on 2.4 or 5 GHz or standard wired Ethernet connections.