Why the Internet of Things needs another ten years

22.01.2016
I first wrote about the hype surrounding the Internet of Things in 2004. Back then, industry analysts predicted that tens of billions of things would be connected to the Internet within five years. It didn’t happen.

Now the hype is back. The 50 billion things that were supposed to be connected by 2010 were merely postponed. The new projected arrival date is 2020.

Don’t get me wrong. The Internet of Things (IoT) is an exciting growth opportunity. (It’s also a misnomer, but that's a discussion for a future post.) The problem is that it is really many different opportunities, each of which will grow at its own pace. A few markets, such as asset tracking and automatic meter reading, have achieved considerable traction. Others, such as connected cars and smart homes, show great promise. And we all look forward to the day when sensors embedded in our clothing detect health problems in their earliest and most treatable stages.

We still have to learn how to find the value in oceans of data. As Francis daCosta suggests in his free Kindle book, Rethinking the Internet of Things, if we deploy enough wind speed, wind direction, barometric pressure and temperature sensors then we may learn to predict precisely where the next tornado will form. The Internet of Things has many unknown unknowns.

Significant progress has been made over the past decade. We now have low-cost sensors, cloud-based services with global reach and big data analytics. What stands in the way of many promising applications (such as supply-chain management) is the dearth of ubiquitous, low-cost connectivity.

Today’s wireless landscape presents a dilemma. Mobile and satellite networks have great coverage, but they weren’t designed for the IoT. New networks were designed from the ground up for IoT, but they don’t yet provide enough coverage.

One way around the problem is to adapt existing mobile phone and satellite networks to IoT needs. Based in North America, Aeris Communications and Kore serve as intermediaries between enterprises requiring broad geographical coverage and mobile operators. The two companies purchase wholesale data service from operators all over the world so that their customers only have to deal with one service provider. Aeris Communications’ platform provides 2G, 3G, and 4G connectivity, device management (such as provisioning and billing), and sensor data analytics. (Aeris’s CTO Syed Z. Hosain has also written a free e-book that provides an excellent introduction to the Internet of Things for business.) Kore pioneered M2M service rate plans, provides unified management of cellular and satellite network services in more than 180 countries, and serves customers using its PRiSMPro platform with redundant data centers in Atlanta and Las Vegas.

Mobile phone and satellite services are a good fit for monitoring critical assets. In some cases, a critical asset (such as an 18-wheeler) may be equipped with multiple sensors and actuators. These can be networked locally using short-range technologies such as ZigBee, Bluetooth or Wi-Fi. The data can be aggregated by a router or gateway and forwarded over the mobile phone or satellite connection.

However, there are many applications that call for communicating intermittently with simple, inexpensive devices that are dispersed around a city or the countryside. These devices could be installed on street lamps, fire hydrants and traffic signals. Or they could be deployed by farmers to monitor soil moisture. They are too scattered to use short-range wireless technologies, and too numerous and cost-sensitive to use satellite or mobile phone service.

Fortunately, a new generation of low power wide area (LPWA) networks are being built that are specifically designed for such IoT applications. These networks can usually cover a city for an order of magnitude lower cost than a mobile phone network, and they are typically optimized to handle many short messages from large numbers of battery-powered devices.

San Diego-based Ingenu (formerly On-Ramp Wireless) “bring[s] connectivity exclusively to machines.” The company says that its proprietary random phase multiple access (RPMA) technology enables wide area connectivity using unlicensed devices in the globally-available 2.4 GHz band. An RPMA access point can cover 30 to 200 square miles, depending on terrain, communicating with devices that can run for years without having to replace their batteries. Ingenu says it can serve devices that individually consume less than $1 of service per month. The company claims 38 private networks operational in 20 countries. Ingenu provides the technology -- operators build their own networks and choose their own business models.

Based in France’s “IoT Valley,” Sigfox competes for similar business using a different approach. The firm sees itself lowering the barriers to Internet of Things connectivity by providing “simple, economical, energy-efficient two-way transmission of small quantities of data over long distances.” Using ultra narrow band (UNB) technology at frequencies below 1 GHz (such as the unlicensed 900 MHz band in the U.S.), Sigfox sells base stations that can cover areas with a radius of a few kilometers (in urban centers) to 50 kilometers (in rural settings). Operators must share a small percentage of their revenue with Sigfox. Sigfox claims 7 million connections in 12 countries (four of which have built out nationwide networks). In addition to familiar applications, creative solutions inspired by Sigfox include a “smart button” that hotels and restaurants use to call taxis and a sensor for detecting when a fire hydrant has been opened.

Based in San Jose, California, Silver Spring Networks uses a mesh network architecture based on the 802.11.15.4g standard (also known as Wi-SUN) supporting speeds up to about 1 mbps. The company focuses on utility networks and smart cities -- what CEO Mike Bell calls “the Internet of critical Things.” One advantage of a mesh network is that it’s easy to extend coverage a little further (in a cellular network, a new base station may be required). Silver Spring claims 22 million devices connected. Customers include the city of Glasgow, Scotland and Oklahoma Gas & Electric.

Given the huge market forecasts, the mobile phone industry is busy developing IoT standards for 4G and 5G. The industry’s answer to radically different architectures (LPWA and mesh) is an assortment of network, protocol, and device enhancements. This includes LTE-M, which will enable devices that use less bandwidth, consume less power and are less expensive. What’s still missing is the motivation to weigh down networks that are expensive to build and operate with things that, individually, generate very little revenue.

Likewise, there are reasons to be skeptical about the role of IoT in 5G. The mobile phone industry has its hands full trying to deliver greater capacity, lower latency, and higher speeds. At some point you have to ask whether it wouldn’t make more sense to build separate networks for IoT.

The Internet of Things will require a range of technologies and business models. It’s good to see more connectivity options from companies with real customers starting to emerge. Because growing the IoT market will continue to take time.

(www.computerworld.com)

Ira Brodsky