Analysis of Electronic Nose Technology for USDA Applications

Problem/Opportunity

The United States Department of Agriculture (USDA) needs an electronic device capable of detecting contraband agricultural products and other prohibited articles in air and sea passenger and cargo shipments. The device or instrument should be capable of supplementing current practices for detection. Electronic nose technology should be capable of detecting fruits, vegetables, and other plant products; meats and animal products; plant pests; pathogens; and other contraband.

An electronic nose, like its biological prototype, consists of an array of chemical sensors, together with a sampling system and a pattern classifier. The collected responses of the array constitute a fingerprint characteristic of the sample. The electronic nose method can be more effective for complex samples than conventional chemical analysis. The USDA requires an electronic nose that is capable of reliably detecting contraband at a port of entry and will yield less than 10% false positives. As a starting point, a review and analysis of currently available and imminently available technology are needed.

Approach

Argonne teamed with the Illinois Institute of Technology (IIT) to perform a literature review and recommend future direction to develop an electronic nose for the USDA.

A wide range of technologies are being identified as e-noses. Argonne and the IIT Research Institute (IITRI) evaluated the available technologies on the basis of detection method and data or signal processing methods. The evaluation addressed the range and type of contraband to be detected, the levels of detection, interferences, and reproducibility.

Results

Argonne identified two serious limitations to electronic nose methods:

• Sensitivity: Most electronic noses have limits of detection in the low ppm range. Limits of detection in the low ppb range or lower are required.
• Matrix Suppression: Many problems involve variable matrices. Isolating a known pattern from a highly variable background will require advanced combination of pattern classification and hardware approaches.

Future Plans

We are currently anticipating additional funding from the USDA to perform laboratory tests upon commercially available electronic noses to determine their utility in USDA applications.