A knowledge-based assistant for pesticide identification

The Pesticide Identification Assistant (PIA) is a knowledge-based system designed to assist an analytical chemist with the identification of pesticide residues in food products using gas-liquid chromatography. The data are in the form of retention times and detector responses for a particular sample. The PIA system aids the analytic chemist in identifying the corresponding residue. The system has four goals: simplify routine analysis; provide interactive training for new analysts; provide an infallible memory supplement for experienced analysts; and provide a means of capturing expertise vulnerable to loss through retirement. A description is given of the problem domain and implementation concerns are addressed, particularly knowledge representation, inference techniques and user interface design. Particular challenges posed by the problem domain, particularly in the area of uncertainty, are considered. Plans for expansion and further research with the system are indicated