Introduction

Sensing system based on single nanomaterial sensors, positioned in a thermal gradient from an integrated asymmetrical heater. The combination of the different responses (like in an electronic nose) and the use of machine learning techniques, allow to achieve good performance both for the classification of gases and the estimation of their concentrations.

Technical features

Resistive sensors composed of a single nanomaterial are placed in a thermal gradient (spatial or temporal) in order to exploit the different working temperature. In practice, sensors which are the same at different temperatures are exploited as different materials in an electronic nose. Less materials and heaters mean smaller size, lower costs and consumption.
The sensors (usually 4-5) and the heater are simultaneously deposited with a single pattern, less than 0.5×0.5 mm2. The responses of the various sensors are combined and are processed by machine learning algorithms.
The combination of nanostructured materials and machine learning allows to achieve a perfect classification (100%) and a very good quantification (average error: 10-15%).
The electronics of the second generation prototype is about half the size of a smartphone, but has great margins for miniaturization.

Possible Applications

• Agrifood (quality monitoring of products, counterfeit analysis…);
• Environmental analysis (air quality both indoor and outdoor);
• Biomedical (breath analysis for early diagnosis);
• Industrial (process monitoring through emitted gases);
• Detection of explosives and dangerous substances.

Advantages

• Simplicity (deposition of only one sensing nanomaterial: contextual deposition of all sensors and heater);
• Inexpensiveness (only one point heater, low consumption);
• Size (less than a squared millimiter);
• Portability and integrability in any device.