As more machines talk to each other directly, humans fumble with policy

Once upon a time, there was a concept in the high-tech industry called pervasive computing. The phrase was used to describe the idea that teeny pieces of communications technology were beginning to pervade virtually every aspect of our life. That virtually everything would be linked, in some form or another, into the Internet. Over time, people ditched the pervasive computing nomenclature and begin referring to the "Internet of things," in which machines talk to other machines with little or any human intervention.

In reality, the Internet of things is already pretty embedded into our daily routine: think of the sensors that log your tolls during your commute or, to cite a more current example, the Internet of things will very much be central to the smart grid. The sensors that keep tab on environmental conditions or other preset factors fall into this category. So do the sensors that are part of things like a Digital Signage prototype being worked on by Intel. The prototype, which was announced earlier this year at the Consumer Electronics Show, could include technology that views who is looking at the screen, assesses their demographics and displays relevant information or advertising based on that analysis. This video gives a sneak peek of what this means.

The goodly researchers at McKinsey Quarterly have just published an essay exploring the state of the "Internet of Things" that explores not just the fact that this happening behind the scenes rapidly but the very real issue that your average business isn't set up to handle this sort of real-time, dynamic analysis. It is not enough to collect this information, you need to be able to act on it.

The McKinsey essay suggests that there are six different application types that will be particularly touched by the Internet of things over time. They are (in no particular order):

• The ability to track behavior (whether it's purchasing patterns or driving patterns), so that pricing can be adjusted to reward certain types of behavior or discourage others. This, obviously, will be big in the supply chain world in the form of things (components, products, etc.) that have radio-frequency identification tags.
• The idea that you can provide situational awareness. So, for example, reporting on environmental conditions such as the weather or ocean currents that might alter logistics decisions. So, you might reroute an airline or truck, for example, if it was more efficient to do so.
• Support for decision analytics. This is the notion that sensors could be used on an oil exploration field, for example, to help guide drilling decisions. Or, that stores could keep tabs on traffic to certain merchandising displays in order to make better inventory decisions. This is also the stuff that's central to digital healthcare monitoring applications.
• Processes of all types could be optimized via the use of sensors that automate what we have traditionally been doing manually, such as where an item is placed on the conveyor belt of an assembly line or whether a kiln or oven is the right temperature.
• I already mentioned applications for the smart grid. One big one is demand response, which is already being used in some data centers to help manage power consumption more rationally.
• Ultimately (and this is out there), McKinsey suggests that we will see the emergence of complex autonomous systems. This is the notion that we would define the appropriate responses for certain sensors and then pretty much get out of the way. The response to certain conditions (say a spill in a manufacturing plant) would be automated, with clean-up robots deployed to address a glitch.

All the principles are fundamental to what we refer to as smarter computing and will be central to pretty much anything you stick the smart adjective in front of today: smart transportation, smart grid, smart cities. That's not because humans can't be smarter than machines. It's because it's smarter to rely on computers and other technology for things that can be automated, so that we human can focus more of our collective brain power on the things that really matter.

This post was originally published on Smartplanet.com