[ Sensing / Connected systems ]

E-Nose

From an unfamiliar air sample to a recognisable pattern: sensing, calibration, and AI-assisted analysis.

My role: Research collaborator and coauthor
Status: Published research
Year: 2026
Methods & tools: Gas sensorsIoTData acquisitionApplied research

THE IDEA, IN MOTION

What could a machine smell?

E-NOSE / 01ILLUSTRATED WORKFLOW

Start with something in the air.

An unfamiliar sample enters a controlled chamber. A pump brings the sample across several gas sensors, just as a laboratory instrument would.

A visual explanation of the research process. Illustrative signals and movement; no measured results are shown.

[ 01 / The context ]

The question.
The approach.

Two air samples can look identical yet contain different chemical mixtures. An electronic nose makes those differences available to study: several sensors respond to a sample, and their combined pattern acts like a fingerprint. This collaborative research develops an adaptable laboratory platform for capturing those responses and exploring how they can be interpreted.

How the work comes together

The 2026 Sensors article, Adaptive E-Nose: Integrating New Gas Sensors for Emerging Applications, documents the adaptable sensing platform. The laboratory setup combines a sensor array, experimental chambers and a data pipeline. The animation extends that workflow into an AI research concept: calibrate the signals, compare their collective fingerprint, and investigate classification models rather than relying on one sensor reading.

Adaptable hardware

Bring heterogeneous, affordable gas sensors into a shared experimental sensing platform.

From sensor to signal

Capture and process multi-sensor responses, supported by a laboratory data-acquisition workflow.

Towards intelligent interpretation

Explore how calibrated response patterns could support machine-learning classification, with uncertainty and experimental validation kept in view.

Two-chamber laboratory setup for the adaptive electronic nose. — The E-Nose laboratory setup, with experimental chambers, sensor hardware and the data-acquisition connections.

[ 02 / Research scope ]

Where the work stands.

The published work describes a laboratory sensing and acquisition platform. AI classification in this visual is a research direction, not a validated identification result. No detection accuracy or operational capability is implied.

Hardware photographs from the Adaptive E-Nose research project.