Robot Ecosystems/Footnotes

Born in 1961. Senior Researcher at Sony Computer Science Laboratories, Inc. Through researching automated audio translation systems development and ultra parallel processing computed artificial intelligence, Mr. KITANO is now attempting to develop a new intelligence systems architecture via molecular/computational biology and RoboCup.

Born in 1960. Senior Research Scientist at Intelligent Systems Division of Electrotechnical Laboratory a.k.a. Densoken (adjunct to the national Agency for Industrial Science and Technology). Engaged in humanoids, interaction, and other perception and action interactive activities intelligence research.

A research field where multiagents (including robots) act in cooperation and competition, learning and acquiring.

Just as computer chess represented, a common theme is established so that numerous researchers can have a common set of problems to address themselves to, encouraging articulation and development of the field. For purposes of unambivalent decision, win/lose competitions are common.

Without tasting the apple, the system attempts to match representations of what has been put in front of it.

A problem which arises in practical applications for intelligent expression systems (a comprehensive expression of data and its means of execution). When a given condition and its transformations provide the input for an intelligent expression system, the system cannot observe only the changes that occur at a given point in time, but must continually account for all aspects of the current surroundings before proceeding to the next.

Based on the thinking that when a given sensory input is obtained the issue of which action to take relies solely upon the condition of the input, not on the previous input or condition of the agent. What is indicated here is the assumption that by more generally considering conditions and surroundings, including the robot's on-going processes, and to a certain extent transitional periods relative to a given situation within the sensory input, operational decisions can occur through a process of learning. This can be inversely stated as that the robot itself should be able to regulate its surroundings, and this is what is essential here.

Emergence. There are several interpretations, however, the definition used here is that "emergence" consists of bottom-up and top-down mechanisms co-defining each other, a process characterized by a mutual synthesis of local and autonomous elements from the lower levels forming a broad organizational order in the upper level, which re-influences lower level operational motivations and border conditions. In living systems, for example, cells (lower level elements) forming (bottom-up) into organs or bodies (higher level broad organizations), while the condition of movements and organs influence the activities of the cells (top-down). Similar relationships can be seen universally throughout multi-layered complex systems.

Isaac ASIMOV's "Three Laws of Robotics", presented in his 1942 short story,

  1. A robot may not injure a human being, or through inaction, allow a human being to come to harm.
  2. A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.
  3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

1930-75. Molecular biologist. Former Nagoya University Professor. As the discoverer of "non-continuative replication," made important contributions to the solution of DNA's replication mechanism. Died from complications due to exposure to atomic radiation.

A method where punishments and rewards are applied to the subject's behavior to reinforce preferred values. Through repetitions thereof, the agent (robot) learns to modify its own actions.

Please see [*8] above.

A method to seek a solution by means of a program written in computer language, treated as chromosomic material (genetic model), applied as a genetic algorithm, and weeded out based on appraisals of its functionality. Genetic algorithm is a process which simulates evolutive behavior. The base algorithm is made to select structures beneficial to it through a sort of "natural selection," and evolve towards an algorithm more adaptable to its environment and the transformations within it.

Where multiple processes maintain an confluent influence throughout their development. Indicating a process in natural science where multiples of seeds interact in their evolution, here it is used to show the learning process occurring between "players" and "coaches" on the robotic playing field.

rightTo Contents Menu