Aggregation of sensory input for robust performance in chemical sensing microsystems

This paper demonstrates the usefulness of aggregating information generated from arrays of chemical sensors for improving the ability to discriminate among target chemicals and their potential interferents. Two types of aggre gation methods are evaluated; the first set do not compress the data, but i ncorporate effects from neighboring sensors into the output of each sensor in an array. The second method does result in compression of data and aggre gates multiple sensor outputs into a single, more robust signal. Methods fo r processing data and aggregating and smoothing outputs from arrays of tin- oxide sensors are comparatively analyzed. Processing parameters studied inc lude those related to simple averaging, linear-weighted averaging, and expo nential smoothing across operating temperature and across type of sensing f ilm in the dimensionality of the array. Aggregation techniques are evaluate d during various stages of both the transient and steady-state response of the array to quantify the early decision-making capability of the array ove r that of a single or small number of unprocessed sensors. Aggregation stra tegies are studied in combination, and results are extracted by quantitativ ely measuring the goodness of clustering for each case. Cluster analysis, i ncluding principal component analysis (PCA), is used to determine which of these processing techniques are most effective. It is shown that aggregation methods, whether they reduce transmission band width or not, improve the performance of a 30-element, tin-oxide heterogene ous sensor array in discriminating among common breath alcohol components ( ethanols), their interferents (acetone, formaldehyde, isopropyl), and a con trast substance (ammonia). Aggregation generates a best-case 42% improvemen t in separability of clusters and 6.25% improvement in the tightness of clu sters. Results are shown that clearly demonstrate the usefulness of aggrega tion in heterogeneous arrays among sensors whose outputs possess an appreci ably degree of correlation (overlapping specificity). (C) 2000 Elsevier Sci ence S.A. All rights reserved.