Search

The competence being investigated is causal modelling, whereby the behavior of a physical system is understood through the creation of an explanation or description of the underlying causal relations. After developing a model of causality, I show how the causal modelling competence can arise from a combination of inductive and deductive inference employing knowledge of the general form of causal relations and of the kinds of causal mechanisms that exist in a domain. The hypotheses generated by the causal modelling system range from purely empirical to more and more strongly justified. Hypotheses are justified by explanations derived from the domain theory and by perceptions which instantiate those explanations. Hypotheses never can be proven because the domain theory is neither complete nor consistent. Causal models which turn out to be inconsistent may be repairable by increasing the resolution of explanation and/or perception. During the causal modelling process, many hypotheses may be partially justified and even leading hypotheses may have only minimal justification. An experiment design capability is proposed whereby the next observation can be deliberately arranged to distinguish several hypotheses or to make particular hypotheses more justified. Experimenting is seen as the active gathering of greater justification for fewer and fewer hypotheses., MIT Artificial Intelligence Laboratory

The basic task of a rover is to move about automonously in an unknown environment. A working rover must have the following three subsystems which interact in various ways: 1) locomotion--the ability to move, 2) perception--the ability to determine the three-dimensional structure of the environment, and 3) navigation--the ability to negotiate the environment. This paper will elucidate the nature of the problem in these areas and survey approaches to solving them while paying attention to real-world issues., MIT Artificial Intelligence Laboratory

Change pervades every moment of our lives. Much of our success in dealing with a constantly changing world is based in common-sense physical reasoning about processes and physical systems. Processes are the way quantities interact over time. Physical systems can be described as a set of quantities and the processes that operate on them. Representations for causality, time, and quantity are needed to fully characterize change in this domain. Several ideas for these representations are examined and synthesized in this paper towards the goal of constructing a framework to support understanding of, reasoning about, and learning how things work., MIT Artificial Intelligence Laboratory