| A knowledge based robotic expert system has been designed and implemented to design, simulate, modify, and execute analytical procedures. A conceptual laboratory model is developed using an object-oriented design technique to simulate a robotic analytical laboratory. The conceptual laboratory model represents the robotic laboratory table as an abstract workbench and the laboratory activities as a set of primitive operations. Chemical properties and chemical reactions are used to construct a deep-reasoning based inference engine, which enables the primitive operations to alter the state of the workbench. By comparing the state of the workbench with that of the real robotic table, the expert system can validate a designed procedure and update the system knowledge base.;A multi-layer structured robot interface protocol is specified and implemented. With a standard robot interface, analytical procedures can be designed and saved at conceptual level using chemistry notation in one laboratory, and delivered to other laboratories and executed with same quality assurance. Analytical procedures for analyzing some of the common elements such as Mn, Cu, Fe, Ni, Co, Mg, Pb, Cr, and Cl have been designed and executed by the robot laboratory successfully.;An object-oriented database management system together with a chemical reaction simulation module is developed to manage chemical information and to simulate chemical reactions. With the chemical information management and the chemical reaction simulation, a user can design an analytical procedure using chemical knowledge at the conceptual level, whereas the expert system validates the designed procedure and controls the robot execution to accomplish the analysis. |
PDF Full Text Request