Expert System for Decision Making and Instructing Nuclear Resonance Fluorescence Cargo Interrogation

Artificial intelligence has been recognized as a highly potential field for automation of cargo scanning process. Specifically, intelligent systems have the capability to direct the process without the interference of a human operator and being able to make decisions. In this paper, an expert system for directing and instructing the cargo inspection, which is performed by means of nuclear resonance fluorescence (NRF), is presented. The empirical knowledge is embedded to the system through fuzzy sets while the inference mechanism is encoded by a set of fuzzy rules. The expert system gets as an input a set of parameters arrived from the NRF processing unit and decides whether the cargo is positive or negative in hazardous materials or no decision can be made. In the last case, the way that the sensing parameters should change, to improve next step detection results, is indicated.