January 6, 2026
Who nose? Mold does
Electronic Nose for Indoor Mold Detection and Identification
Sniffing out toxic mold — homeowners want DIY, nerds want graphene, skeptics want ions
TLDR: A new “electronic nose” detects and identifies indoor mold with near-perfect accuracy. Commenters split between wanting a DIY whole-house scanner, arguing for graphene upgrades, and touting negative-ion hacks, while supporters say it could empower sick homeowners with faster, cheaper proof their air is making them ill.
The lab just gave us a gadget that literally “smells” your home: an electronic nose that can sniff out mold and even tell which kind it is. Traditional tests take days and cost big money, and mold-detection dogs can’t tell species apart. This prototype uses tin-oxide nanowires and smart pattern-matching math to hit a reported 98% accuracy identifying the two usual suspects—black mold Stachybotrys and Chaetomium. Cue the comments: DIY warriors like andsoitis want a whole-house sweep—“Can I scan every room myself?”—while hardware nerds like westurner immediately yell “add graphene!” to boost sensitivity at room temp. Patient advocates cheer that this could finally stop sick folks being dismissed or “gaslit.” Then the curveball: HocusLocus suggests blasting “negative ions” and just wiping walls—sending the thread into eye-roll city. The meme factory spins up a “Roomba with a nose” and a “dog vs robot sniff-off,” while pragmatists ask when this becomes a real product and how much it’ll cost. Translation for non-nerds: the e-nose catches the smells mold gives off, compares them to known patterns, and even flags odd odors. The community is split between wanting a cheap home gadget now, upgrading it with shiny materials, or side-stepping tech entirely with questionable hacks.
Key Points
- •The study develops an electronic nose using UV-activated, VLS-grown SnO2 nanowire chemiresistive sensors to detect indoor molds.
- •Two common indoor molds, Stachybotrys chartarum and Chaetomium globosum, were tested on two substrates relevant to water-damaged buildings.
- •Machine learning via Linear Discriminant Analysis (LDA) was used for odor pattern classification; novelty detection was provided by decision boundaries.
- •Conventional LDA gave mediocre results, but improved LDA versions achieved an average F1-score of 98.37%.
- •Findings indicate the e-nose can both detect and identify mold genera, enabling faster, more objective, and cost-effective indoor air quality monitoring compared with traditional methods.