Two years ago, the Chrysler corporation completely gutted its Windsor, Ontario, car assembly plant and within six weeks had installed an entirely new factory inside the building. It was a marvel of engineering. When it came time to go to work, a whole new work force marched onto the assembly line. There on opening day was a crew of 150 industrial robots. Industrial robots don’t look anything like the androids from sci-fi books and movies. They don’t act like the evil Daleks or a fusspot C-3P0. If anything, the industrial robots toiling on the Chrysler line resemble elegant swans or baby brontosauruses with their fat, squat bodies, long arched necks and small heads. An industrial robot is essentially a long manipulator arm that holds tools such as welding guns or motorized screwdrivers or grippers for picking up objects.
The robots working at Chrysler and in numerous other modern factories are extremely adept at performing highly specialized tasks – one robot may spray paint car parts while another does spots welds while another pours radioactive chemicals. Robots are ideal workers: they never get bored and they work around the clock. What’s even more important, they’re flexible. By altering its
The word robot is Slav in origin and is related to the words for work and worker. Robots first appeared in a play, Rossum’s Universal Robots, written in 1920 by the Czech playwright, Karel Capek. The play tells of an engineer who designs man-like machines that have no human weakness and become immensely popular. However, when the robots are used for war they rebel against their human masters. Though industrial robots do dull, dehumanizing work, they are nevertheless a delight to watch as they crane their long necks, swivel their heads and poke about the area where they work. They satisfy “that vague longing to see the human body reflected in a machine, to see a living function translated into mechanical parts”, as one writer has said.
Just as much fun are the numerous “personal” robots now on the market, the most popular of which is HERO, manufactured by Heathkit. Looking like a plastic step-stool on wheels, HERO can lift objects with its one clawed arm and utter computer-synthesized speech. There’s Hubot, too, which comes with a television screen face, flashing lights and a computer keyboard that pulls out from its stomach. Hubot moves at a pace of 30 cm per second and can function as a burglar alarm and a wake up service. Several years ago, the swank department store Neiman-Marcus sold a robot pet, named Wires. When you boil all the feathers out of the hype, HERO, Hubot, Wires et. al. are really just super toys.
You may dream of living like a slothful sultan surrounded by a coterie of metal maids, but any further automation in your home will instead include things like lights that switch on automatically when the natural light dims or carpets with permanent suction systems built into them. One of the earliest attempts at a robot design was a machine, nicknamed Shakey by its inventor because it was so wobbly on its feet. Today, poor Shakey is a rusting pile of metal sitting in the corner of a California laboratory. Robot engineers have since realized that the greater challenge is not in putting together the nuts and bolts, but rather in devising the lists of instructions – the “software – that tell robots what to do”. Software has indeed become increasingly sophisticated year by year. The Canadian weather service now employs a program called METEO which translates weather reports from English to French.
There are computer programs that diagnose medical ailments and locate valuable ore deposits. Still other computer programs play and win at chess, checkers and go. As a results, robots are undoubtedly getting “smarter”. The Diffracto company in Windsor is one of the world’s leading designers and makers of machine vision. A robot outfitted with Diffracto “eyes” can find a part, distinguish it from another part and even examine it for flaws. Diffracto is now working on a tomato sorter which examines colour, looking for no-red – i.e. unripe – tomatoes as they roll past its TV camera eye. When an unripe tomato is spotted, a computer directs a robot arm to pick out the pale fruit. Another Diffracto system helps the space shuttle’s Canadarm pick up satellites from space. This sensor looks for reflections on a satellites gleaming surface and can determine the position and speed of the satellite as it whirls through the sky.
It tells the astronaut when the satellite is in the right position to be snatched up by the space arm. The biggest challenge in robotics today is making software that can help robots find their way around a complex and chaotic world. Seemingly sophisticated tasks such as robots do in the factories can often be relatively easy to program, while the ordinary, everyday things people do – walking, reading a letter, planning a trip to the grocery store – turn out to be incredibly difficult. The day has still to come when a computer program can do anything more than a highly specialized and very orderly task. The trouble with having a robot in the house for example, is that life there is so unpredictable, as it is everywhere else outside the assembly line. In a house, chairs get moved around, there is invariably some clutter on the floor, kids and pets are always running around.
Robots work efficiently on the assembly line where there is no variation, but they are not good at improvisation. Robots are disco, not jazz. The irony in having a robot housekeeper is that you would have to keep your house perfectly tidy with every item in the same place all the time so that your metal maid could get around. Many of the computer scientists who are attempting to make robots brighter are said to working in the field of Artificial Intelligence, or AI. These researchers face a huge dilemma because there is no real consensus as to what intelligence is.
Many in AI hold the view that the human mind works according to a set of formal rules. They believe that the mind is a clockwork mechanism and that human judgement is simply calculation. Once these formal rules of thought can be discovered, they will simply be applied to machines. On the other hand, there are those critics of AI who contend that thought is intuition, insight, inspiration. Human consciousness is a stream in which ideas bubble up from the bottom or jump into the air like fish. This debate over intelligence and mind is, of course, one that has gone on for thousands of years. Perhaps the outcome of the “robolution” will be to make us that much wiser.