Presentation Details
| A Human-Nose-Inspired Biomimetic Electronic Nose Kobi Snitz1, Tali Weiss1, Arbel Arad1, Danielle Honigstein1, Aharon Weissbrod1, Nir Harel2, Noam Sobel1. 1Weizmann Institute of Science, Rehovot, Israel.2 Bezalel Academy of Art and Design, Jerusalem, Israel |
Abstract
Electronic noses (eNoses) have been used in a variety of applications from detection and classification of odorants to predicting their perceptual semantic descriptions. Extensive efforts have been directed at sensor technology, but despite their names, the design and function of electronic noses has little to do with that of the biological system. We built a human-nose–inspired biomimetic eNose to test the functional value of specific nasal features and sampling strategies. Using high-resolution 3D printing, we reproduced a CT-derived model of a human female nasal cavity. Mass-flow controllers (MKS) at the pharynx generated realistic breathing waveforms, and sensors were positioned at the olfactory cleft—the native location of the olfactory epithelium. The platform can host diverse sensing technologies; here we first tested two Airsense PEN 3 units (metal-oxide array, 10 coatings each), one per nostril, set to sample at 10Hz. To mimic realistic sampling, odors were not provided by olfactometer or autosampler, but were rather sniffed from wide-mouth jars, mimicking a typical human olfactory experience. In an initial experiment we asked whether olfaction benefits from treating the previous inhalation as a running, dynamic baseline. We classified naturalistic odorants drawn from a list of the most commonly encountered smells (e.g., pine needles, fresh fruit, chocolate, peanut butter). Implementing a running baseline improved classification accuracy by 45% (p <0.0001). This result suggests a practical strategy for coping with dynamic odor environments that immediately boosted eNose performance. Additional biomimetic manipulations will be presented.
No part of this publication may be reproduced, distributed, or transmitted in any form or by any means, including photocopying, recording, or other electronic or mechanical methods, without the prior written permission of the author.
No part of this publication may be reproduced, distributed, or transmitted in any form or by any means, including photocopying, recording, or other electronic or mechanical methods, without the prior written permission of the author.
Content Locked. Log into a registered attendee account to access this presentation.