Earlier walking robots had a more serious side: to discover more about the human brain. Shakey, for example, emerged from the Stanford Research Institute (now SRIs artificial intelligence centre) in 1
Earlier walking robots had a more serious side: to discover more about the human brain. Shakey, for example, emerged from the Stanford Research Institute (now SRIs artificial intelligence centre) in 1972 and differed from Elektro in key respects. It was mobile, spatially aware and could move around without bumping into things. Shakeys eyes consisted of a TV camera, and his sense of touch was provided by a series of bump sensors, both of which were connected to his robot brain: DEC PDP-10 and PDP-15 computers. Shakeys software also allowed him to turn, plan routes and to remember where hed been.
Shakeys influence can be seen today in robots such as Carnegie Mellon Universitys tour guide, Minerva. The robot roams autonomously through the daily crowds at the Smithsonians National Museum of American history and perceives her surroundings through cameras, laser range-finders and ultrasonic sensors: this information is interpreted by on-board computers. Minerva actively approaches visitors to the museum, offers tours and leads them from exhibit to exhibit. When shes happy, she smiles and sings at people; however, if her way becomes blocked too often, she can, like us, become frustrated and will frown and honk her horn.
Shakey was closely followed in 1974 by a life-size, humanoid robot called WABOT-1, created by Tokyos Waseda University (WAseda roBOT). WABOT-1 had limb control, a vision system and a conversation system, and was capable of holding a conversation in Japanese. It could also measure distances and directions to objects using external receptors, artificial ears and eyes and an artificial mouth. By 1984, Waseda had produced WABOT-2, which read music and played an electronic organ at the Tsukuba Science Expo.
These robots were the product of a new field of research: Artificial Intelligence (AI). Alan M Turing inspired the field of AI in 1950 when he published his, now famous paper Computing Machinery and Intelligence, which posed the question: can machines think? In the paper, he describes a way in which humans can test the level of artificial intelligence in a machine, known today as the Turing Test. In its most basic form, a human judge sits at a computer posing written questions. The judge must then decide whether theyre communicating with another human being or an AI program.
One of the early products of his paper was the now equally famous AI Laboratory at the Massachusetts Institute of Technology. The Lab is responsible for Cog, an ongoing project to create a mechanical torso thats as human-like as possible. Cog already has an artificial skeleton comprised of more than 20 actuated joints, and twice as many sensors, ranging from torque sensors on motors to the four cameras composing the eyes. The aim is to create an artificial nervous system for Cog, so that it can recognise which of the impulses from its brain controls functions such as moving an arm or leg. Once Cog has learnt this, its hoped it will be able to move its limbs based on the desired effect.