Abstract
Operations automation is automation that replaces, wholly or in part, operational activities currently carried out by human controllers in complex systems. It is intended to be neither ‘black-box’ nor ‘human-tended’ automation, but rather automation that functions independent of human control and yet still facilitates its inspection and repair as necessary. This paper describes research to develop an apprenticeship approach to developing a knowledge base to support such automation. The result is both a human-centered automation approach and a software architecture, Apprentice, to support this approach. Apprentice enables human operators to create the knowledge base for operations automation by performing their normal control activities. Apprentice watches and compares them with those specified in the knowledge base, noting discrepancies between the knowledge base and operator activities. Graphical knowledge base editing tools are then used—-by the domain practitioners—-to modify, refine, or extend the knowledge base as required to account for the detected discrepancies.
Get full access to this article
View all access options for this article.