Augmentation & Robotics

This page is no longer maintained.
Some links are broken.

Human Interfaces & Augmented Reality

Applications you might not have considered


A Coin With Two Sides

The other side of the robotics coin is called "augmented reality". AR is about providing people with more (or more relevant) information than what their senses naturally produce, generally in some highly specific context. An AR system might be used to guide a surgeon's hand, coach a repairman, superimpose radar data over a pilot's view, guide a taxi driver through an unfamiliar neighborhood, help a fireman locate the safest path through a burning building, or help an unskilled worker accomplish a complicated assembly task.

Why Robots?

Despite recent developments with both Honda and Sony showing seemingly production-ready humanoid robots, practically speaking, and especially in the near term, robotics is primarily about replacing materials, energy, and effort (time, attention, and frustration) with machine intelligence.

Take a look at modern agriculture. Fields are tilled, planted, harvested and tilled again by giant machines that are best adapted to uniform crops grown on uniform soil. The reason? The existence of such machines has made older, slower, more attentive methods uneconomic, not because it costs more to run smaller machinery longer, but because it means more man-hours, which either adds to the cost in the case of hired labor, or limits the amount of land one can manage, for those who do the work themselves. What gets lost, along with terraces, waterways, and hedgerows, which encumber the huge machines, is something that used to be called "the footsteps of the farmer" -- another way of saying attention to detail. With a greater number of smaller, slower machines that are able to operate independently, the amount of attention given to each patch of ground can be dramatically increased, while at the same time much of the fuel used can be replaced with less concentrated, renewable sources -- the sun and the wind.

With six billion people on the planet, it might be difficult to argue that there aren't enough to go around, and it might seem like we really shouldn't be worrying ourselves about automating tasks people could do while many still remain unemployed. But what if those tasks aren't worth enough, in the market, to subsidize hiring people to do them? Robots will eventually be able to fill the gap between what might just as well be left undone and what the market deems important enough to pay people to do.

What is Robotics?

Robotics includes a lot more than machines that move around like and interact with humans. Most robots are task-specific and possess processing power, but, for the present, no intelligence beyond what's codified into their programming. Some are quite simple, mechanically, and not particularly expensive to build. (In fact some, called "bots", aren't mechanical at all, but exist purely as software.)

A robot is essentially a system with some means for gathering information about its environment, which it then utilizes to modify its behavior. Robots typically have some sort of sensory apparatus, and one or more processors that 'make sense of' the sensory input and run programs which are contingent upon it. The term for the general study of this kind of arrangement, including natural as well as artificial systems, is "cybernetics".

Emerging Technology, Economies of Scale, and Synergistic Returns

Computers just keep getting more powerful, cheaper, or both. Capabilities, like machine vision, which only a few years ago were the stuff of well financed research labs, are now within the reach of mass market computers, and will likely soon become available as plug-in parts. As the power/cost ratio of robotic systems increases, the range of applications for which they are cost effective broadens, increasing the size of the market and driving down the per unit cost of development, lowering the price, which further broadens the market.

With a larger market, it becomes reasonable to mass produce parts which would otherwise be custom, thus dramatically lowering their cost and still further broadening the market.

At some point the process of manufacturing itself becomes streamlined, if not entirely automated, making many niche applications practical. This would be about the time that the whole scenario goes ballistic...or, more realistically, becomes bound not by technology nor by the costs of production, but by more fundamental socio-economic constraints.

Following this logic is a fairly simple matter, what isn't so obvious is that it holds the potential for a qualitative transformation of the human environment, for the better. If robotics fails to live up to that potential, it won't because the technology doesn't lend itself to that end (it does), but because the inertia of wasteful, destructive habits was too great to overcome.