Artificial intelligence and robotics - visions and realities

Artificial intelligence (AI) and robotics are currently two of the fastest developing fields of research. The idea of constructing an artificial mechanical being has been actively pursued for centuries. One of the long-term aims of current research is to create a robot football team capable of beating the human football world champions by the year 2050.

The term artificial intelligence originated in the USA in 1956 with efforts to simulate human intelligence by means of computer programs. Attempts were made in the 1960s to develop programs that could mimic human thought processes in solving problems. Expert systems were supposed to make computerised know-how available to all. One result of these research efforts is the realisation that human knowledge cannot simply be transferred to logical systems of thought.

These days, research into AI focuses on concrete areas which can be formalised and is achieving remarkable success. Chess programs can now only be beaten by the world's top players. Artificial neuronal networks are the first hardware and software structures with a limited learning capability. Pattern recognition, for instance of images or the sound patterns of human language, makes communication with machines an increasingly simple affair. It also enables technical systems to find their bearings in an environment geared towards humans.
 

Robotics is a kind of applied AI, and makes automated actions flexible. Sensors gather information on the environment, while actors such as electrical or hydraulic drives interpret the intelligently controlled movements. Artificial beings created in the computer - avatars - now face us as "virtual robots". However, they are still not sophisticated enough to pass the Turing Test devised by British AI pioneer Alan Turing in 1950 as a benchmark for artificial intelligence.

[Overview]

[Next]