One of the challenges in tracking new technologies is understanding when they transition from hype (usually called "vision") to reality. I'm not referring to the well-known "crossing the chasm" between early users of technology and mainstream markets. I'm talking about a much earlier transition, from dream to real technology. The Internet of Things ("IoT") is nearing that transition point right now, and the battle is looming between the mobile and chip companies that are heavily researching the area, including Qualcomm (QCOM), Intel (INTC), Cisco (CSCO), Google (GOOG), Microsoft (MSFT), Apple (AAPL), ZTE (OTC:ZTCOF), IBM (IBM), General Electric (GE) and more. Before talking about the companies, let's take a look at the Internet of Things vision itself.
The Internet of Things refers to having our real-world physical environments connected to the Internet and accessible from computers and smartphones. Like the Internet itself, the Internet of Things is envisioned doing many different things: finding a lost set of keys by wirelessly asking the key chain where it is, turning on your air conditioning from the car as you're on the way home, getting an alert on your phone when your washing machine is leaking, being notified when your teenager drives the car into a dangerous neighborhood, reading on your smartphone about the materials used in the dress you're trying on in the store, seeing the day's weather forecast on your bathroom mirror in the morning, finding out how many eggs you have in your refrigerator when you're at the supermarket, and much more.
The dream of the Internet of Things isn't new. From the "Jetsons" to "Star Trek", TV shows and movies have long envisioned a world in which our technology can help us control our environment. More recently, these dreams have moved from science fiction to corporate visions of the future.
Even thirty years ago, engineers were playing with technology that would now be called the Internet of Things. In the early 1980s, in my second year of college at Carnegie Mellon, I first used the famous networked Coke machine, a Coke machine that could be checked online as to whether it was loaded and whether the Coke in the machine was cold. This was before Web browsers and indeed before mass-market e-mail, and it worked -- department folks could check that there was cold Coke before walking up or down steps to the machine. But it was built with the engineering equivalent of chewing gum and toothpicks.
The recent proliferation of wireless and cellular technologies is making the Internet of Things much more realizable. For a few extra dollars in component costs, things of all sizes can connect to the Internet. And most devices nowadays are already using computer chips for control.
But big challenges still remain. If my "things" are connected to my Internet router at home, how can I access them from outside my home, in such a way that the rest of the world can't access them? If my smartphone, my connected thermometer and my connected thermostat are all made by different companies, how can they work together to enable me to check my house's temperature and activate the air conditioner? Will I need a different app for my coffee machine, my bread maker, my television and my air conditioner, or can they all work together with a single app? And how can a variety of sensors, sensing things like water behind a washing machine or the door opening on the liquor cabinet, have their signals combined in a way that can send appropriate notifications or take desired actions?
The challenges get even bigger when we consider the Internet of Things on a metropolitan scale. How can data from speed sensors on highways be organized when there are thousands of such sensors around a city? How can stoplights be connected without opening them up to hackers? How can thousands of connected "things," from street lights to gunshot sensors, be used effectively without being overwhelmed by the quantity of data?
In the past few years there has been an extraordinary amount of concrete technology research addressing all of these challenges. A recent report overviews research by over a dozen major companies, including mobile, technology, software, network and computer chip companies. The Internet of Things is promising to be one of the next battlefields between these companies. Most importantly, each company is researching the Internet of Things in a manner that fits their company's own focus and strength.
In Cisco's vision, the significant technology exists in the network. It envisions over 50 billion "intelligent things" connected to the Internet by 2020. Cisco's research explores how "minimalistic connected objects" can be connected to the network in a manner in which the objects can be very simple. The brains will exist in the network itself, in the form of "distributed intelligence agents" that organize the network of simple objects and manage the flow of information between the simple objects and Internet servers. These are only two examples of Cisco research in the Internet of Things, which show Cisco's agenda of building the IoT's smarts into the network itself. In June, 2013, Cisco launched a new business unit that consolidated IoT-related R&D.
Not surprisingly, Qualcomm's vision puts the emphasis on the things themselves, specifically how things like coffee machines can be connected to the Internet and accessible on smartphones and computers. Qualcomm's R&D, including its Alljoyn alliance, is looking at putting the smarts into the "things" themselves, which of course would require the connected objects having computer chips. Its research is focused on how different "things" can offer information and services that can be accessed by other devices and by permitted Internet users. Similarly, Intel's research in Internet of Things is looking at how computing power can be added throughout the Internet of Things.
Mobile companies are also looking at the Internet of Things in ways that fit their own strengths. Google is developing a platform in which its Android operating system can interact with hardware devices. It is also engaged in advanced research on how sensor data can be collected and made available by a centralized system, much as Google's search engine does for web pages. This includes how its cloud computing platforms can be used to process Internet of Things data.
Microsoft is researching Internet of Things in a numbers of directions, some based on its Xbox gaming device, some based on media sharing, and some based on mobile. Apple is researching "personal items networks" that connect a wide range of nearby personal things to the user's iPhone, along with a novel proximity sensor for sensing where accessories or tags are relative to an iPhone. Nokia (NOK) has prototyped tags that can be connected to key chains and other small objects, enabling them to be tracked by Nokia phones. It too has significant research underlying the technology used in its tags. ZTE is engaged in a comprehensive research program on Internet of Things, supporting mobile applications such as healthcare and travel, and have also made business alliances related to Internet of Things.
Meanwhile, IBM is researching Internet of Things in terms of how servers can collect and analyze data from sensors and other things to generate information that can be used effectively by users and by other programs. General Electric and Bosch are working heavily on how intelligence and connectivity can be added to a wide variety of appliances.
Meanwhile, this report also details more than twenty-five start-up companies that are developing Internet of Things devices, including tags to enable phones to find keys and purses, thermostats that enable you to start your air conditioning from your car, and sensors that tell you when your washing machine is leaking or your locked cabinet is opened. Start-ups are also working on "hubs," devices that connect to, and manage, a variety of sensors and other things.
In the short term, it appears that a wide range of connected "things" are going to be on the market by the end of 2013 and the first quarter of 2014. Early products are likely to be built with whatever can work, as was the CMU Coke machine, but soon companies will centralize on common hardware and software components, letting manufacturers focus on the distinctive aspects of each product.
In the longer term, infrastructure like that being researched by the major network and technology companies will be needed to manage the millions of devices that will be connected. Today's research and initial products are only the start of the battle. Over the upcoming years, we'll see not only the race to bring products and platforms to market, but also the race to define the approach to IoT. Will the key technologies be in the network, in the devices, in smartphones, in web servers, or all of the above?