by Chris Parlato
"Enabling products to tap into the experience of other objects on the network, to learn from a collective history of interactions."
According to one prevailing forecast of technologists (and venture capitalists), within 5- 10 years, the mundane physical objects that surround us
in our daily lives-from the drinking cup to the trash can, to the bathroom mirror-will no longer be silent, static, and isolated. Gathering,
processing, and displaying data from arrays of cheap embedded sensors, these simple objects will be endowed with a rudimentary form of intelligence
that enables them to respond, react, and communicate.
The "smart" object has already become a familiar refrain in the popular lexicon. We have smart phones, smart cars, smart refrigerators, smart products of all sizes, shapes and price points. "Smart" has come to refer to inanimate objects that possess any of a range of dynamic qualities with which they are not traditionally associated, including autonomy, adaptability, reactivity, and expression. With the emergence of inexpensive wireless communication technologies, "Smart" also increasingly means connected-connected to control systems, to sensor arrays, and to the collective intelligence of the internet.
The emerging connectivity of consumer products has powerful implications. Collective usage histories and real-time trends help the product to anticipate and respond to the changing needs of its user faster and more sensitively. New efficiencies become possible as products adapt their use of resources to balance system-wide deficits or surpluses. Finally, dynamic variations produced by the response of a single product to the behavior of the network begin to instill these products with a rudimentary form of animation, manifested through the life-like qualities of serendipity, surprise, and variation.
If machines could learn from the experience of their peers, they would be able to solve problems faster and adapt to new circumstances more
easily. That's the premise behind RoboEarth, an initiative to develop a "World Wide Web for robots,"-a standardized, common language of protocols
and a more modular design of robotic systems that would allow all robots of all kinds, from Roombas to factory welders, to share information. Using
a networked information database RoboEarth system would collect and store information about object recognition, navigation, and task performance,
and transmits this data to robots linked to the network.
Serious hurdles remain in the implementation of universal standards, but there are promising opportunities to apply such a system at the scale of a single product or appliance line.
How might connected smart products learn from each other?
In a rare instance of cooperative foresight, technologists, government planners, corporations, and theorists have in recent years embraced a big
idea: that integration of energy utilities into an interdependent network, fed by real-time consumption and production data and controlled at multiple
scales by regional command centers, will produce significant gains in efficiency at both the urban and household scale.
In this type of Smart Grid system, energy rates are constantly in flux as regulated market adjusts to consumption and production volumes. Miele, a German home appliance manufacturer, responded to this trend with a line of washer-dryers designed to monitor real-time utility rates via the internet, and to operate automatically only when the electricity supply is the cheapest.
How might real-time data from the internet make smart devices more efficient?
Recommendation algorithms, from Netflix to Amazon to Youtube, crunch statistical data from millions of user interactions to make targeted
suggestions-what movie you will probably like, or what products you are likely to be interested in. In general, the more you use these services,
the more they know about you and the better they get at meeting your needs and interests. With the advent of sensor-embedded "smart" products
that generate and share a large amount of high-resolution data to connected networks, the logic of statistical recommendation is migrating from
the PC and mobile environment into a range of connected household appliances and products.
The ThinQ refrigerator, part of a suite of connected appliances from LG, helps users keep track of the shelf location and expiration date of foods stored within it. The ThinQ fridge also offers links to pre-programmed and online recipes based on the available food in the fridge (similar to web services like Supercook and RecipeMatcher). More sophisticated social components (Most Viewed, Top Rated, Recommended Based on Previous Choices, etc.) are not far behind.
How might crowdsourced recommendation systems be integrated into smart products and appliances?
Ambient awareness is a term used by social scientists to describe a new form of "peripheral social awareness propagated from relatively constant
contact with one's friends and colleagues online". The concept is illustrated most essentially by the Twitter feed, a continuous stream of thoughts
and activities that while rarely read in its entirety, provides a reassuring sense of low-intensity social presence in the background of a users'
Ambient Orb-a "data clock" that translates select information feeds, including social network data, from the internet into the chromatic variation of an LED fixture—is an example of how dynamic, real-time information can be communicated through non-screen-based devices. The device, particularly when connected to a social feed, also suggests how ambient social awareness, with its emotional benefits, might become a feature of a variety of consumer products-whether tied specifically to the user's online relationships, or extended to the product's real-time usage network.
How might connected smart products create emotional benefit through forms of ambient social presence?