Googles gamble on thingternet

Written by Pratik Kanjilal | Updated: Jan 23 2014, 08:31am hrs
Buying out connected home device maker Nest for $3.2 billion in hard cash, Google has put its money on the Internet of Things or Thingternet. It has paid a huge premium on a company valued at $2 billion, and its a sound gamble. Couple the IoT with the rise of big data and peer-to-peer, always-on, always-connected computing over IPv6 and you have the makings of a boom which is almost already happening. Googles expression of interest can only accelerate what was upon us anyway. In short, if you know people who are building fiddly little machines which have RFID or their own IP addresses and can talk to each other across the internet, invest hand over fist and become incredibly rich in about five years. The word is that such devices, ranging from coffee makers to baggage trackers, will number in the billions by 2020. Google is buying in early to ensure that most of them run Android. It is a huge market, ranging from just-in-time global inventory management to refrigerators which know precisely how much beer they contain and can order electronically as needed.

With Nest, Google has chosen to invest first in the intelligent home. Back in 2011, executive chairman Eric Schmidt had indicated an interest in home automation, a theme that reappeared in his 2013 book, The New Digital Age. The idea of the intelligent home, whose Venetian blinds adjust automatically to ambient light, and whose climate control can be triggered by residents well before they get home, has been the stuff of science fiction for half a century. But it was traditionally conceived as a centrally controlled facility on the lines of the client-server model. Schmidt introduced a crucial change: in his automated home, devices talk to each other and to external servicesdirectly, not through a central server, not with human oversight.

Google had invested first in the Open Automotive Alliance and Sebastian Thruns driverless car. Now, with the Nest acquisition, it is taking a position in the homes and offices which that car will commute between. The companys flagship product is a net-connected thermostat which saves the user the bother of reading the manual. It programs itself, based on how the user changes temperatures through the day, and on temperature and humidity data picked up by its own sensors. Since it is net-connected, it can be controlled remotely by a phone app. But in the future, if the car and the thermostat are running the same operating system, based on GPS data, they would be able to conspire to turn off the climate control when the car drives away and turn it on again when it heads home, without user intervention.

Thats a whole new consumer market, and the operation of millions of units of such connected devices would generate vast volumes of data, opening up another new market for analysis. A data set as simple as the mean preferred home and office temperature and humidity across regions and nations would be enormously valuable for governments, architects and planners in an energy-deficit world. Clothing manufacturers could use it to design home apparel. Power corporations would anticipate changes in demand on the fly and do better load balancing. And, of course, efficient climate control equipment would be built.

Multiple uses can evolve for a simple, unidimensional data set. If it is correlated with other sets derived from devices with other sensory capabilities, the possibilities would multiply exponentially. If there is an intelligent refrigerator in the house which can sense the rate of depletion of its contents and report to retailer networks, markets would be better able to anticipate seasonal changes in consumption patterns and manage inventories accordingly. These are obvious, traditional uses for data, just half a step ahead of the eternal verities, such as, consumers buy more colas in summer and more meat products in winter. But the operation of big data architectures on the Internet of Things would probably reveal patterns of human behaviour that are unknown and even unsuspected. The discovery of such patterns could fuel the next wave of marketing and product design, opening new niches and possibly even industries feeding needs whose existence we are unaware of.

That is precisely why privacy concerns are already being voiced about Googles plans for Nest, even before its offices relocate to Palo Alto. Google in the home is a step uncomfortably closer than Google in the car, and it already has its foot in the door. The company subsidises users in exchange for monetising their data. Its machines read your Gmail and track your Google searches to serve you ads more accurately. It is more likely to influence consumer choice than any PR firm and with multiplexed data sourced from homes, Google can be even more effective. Indeed, the real challenge for Google could be failsafe privacy assurance, not the technology of the Internet of Things itself.