Intelligent gas monitoring for laboratory and technological environments
VentiSENSE is a web-based demonstrator of an intelligent gas-monitoring system integrating optical sensing, chemical principles, data analysis, artificial intelligence and digital information delivery through QR technology. It transforms an otherwise imperceptible change in the environment into meaningful, interpretable information for the user.
We are an interdisciplinary student research team working at the intersection of chemistry, sensor technology, data science, artificial intelligence and web technologies. Our objective is to explore how these disciplines can be integrated to address a genuine challenge in environmental and occupational safety.
VentiSENSE is not conceived merely as a device that produces a measurement. We are developing a system capable of detecting a change, contextualising the resulting data, interpreting its significance and communicating the outcome in a form that is accessible to the end user.
VentiSENSE is a digital demonstrator of an intelligent gas-monitoring architecture for laboratory and technological environments.
Many gases are imperceptible to human senses, and changes in their concentration may not be immediately apparent without appropriate instrumentation. In laboratories, pharmaceutical plants and technological rooms, measurement alone is not sufficient – data must be contextualised and communicated so the user understands what has changed and why it may matter.
A model polymer-based fluorescent sensor changes its optical signal in the presence of a target gas. An optical readout turns it into data, a web application with an analytical (AI) module assigns a risk category, and the user receives an alert and a recommendation – also via the sensor’s QR code.
Effective environmental monitoring depends not only on acquiring reliable measurements, but on converting them into information that can be understood and acted upon. We demonstrate how interdisciplinary research can turn a chemical signal into a coherent digital information system.
Detection, analysis, interpretation and communication form one continuous information pathway.
A change occurs in the concentration or presence of a selected gas in the surrounding environment.
A model polymer-based fluorescent sensor responds to the target analyte with a measurable change in its optical signal.
The optical response becomes digital data, analysed with measurement time and environmental conditions.
The analytical layer assigns a risk category within the demonstrator’s model and generates a procedural recommendation.
Each sensor has a unique ID and QR code linking the physical measurement point to its digital record.
Result, risk status, history, charts and recommendation are available in the web interface.
The signal change is expressed as
ΔI = (I₀ − I) / I₀ × 100%
where I₀ is the fluorescence intensity before exposure and I the intensity after exposure. The analytical module then considers the gas, the rate of change, exposure time, temperature, humidity and recent history.
| ΔI | Category | Action |
|---|
Concept visualisations prepared by the VentiSENSE team. They illustrate the intended system – they are not photographs of a validated device.
Each sensor may be assigned a unique identifier, for example NH3-01-LAB-A. Scanning the associated QR code gives direct access to the sensor’s digital profile, which can contain:
Scan this code with your phone to open the card of NH3-01-LAB-A during the simulated NH₃ scenario – no physical sensor needed.
Open the sensor cardRather than presenting raw sensor output, the analytical module provides contextual interpretation of the available data.
A transparent rule-based analytical module classifies the risk and explains every decision. In parallel, machine-learning models (Random Forest, Logistic Regression, Decision Tree) trained in Python on our dataset are compared with it – including the confusion matrix.
At the present stage the system relies on modelled and synthetic datasets. Incorporating experimental laboratory data is the next stage of development and validation.
See the model reportA conversational assistant is built into every page. It can:
For each gas: where it matters, why it is relevant and which scenario the demonstrator simulates.
The presence of a gas may be associated with an emission, a leakage, a technological process or inadequate ventilation. Its interpretation depends on the identity and concentration of the gas and on the conditions of the measurement.
That is why VentiSENSE is not designed around detection alone: data is structured, analysed and contextualised before it is presented in an intelligible form.
At a hypothetical laboratory workstation, a user scans the QR code of sensor NH3-01-LAB-A. The system shows a change in the NH₃-related signal, processes it with the risk model and assigns a Caution (moderate risk) category with the recommendation: ventilate the area, repeat the measurement in 2 minutes and inform the laboratory supervisor.
Run this scenario liveWorkstation-level monitoring with QR cards for students and staff.
Process areas, utilities and technological rooms.
Structured monitoring and interpretation of environmental data.
Teaching rooms and storage areas requiring indoor air awareness.
VentiSENSE originally explored the detection of airborne allergen proteins. We shifted the analyte to volatile laboratory gases while keeping the VentiSENSE architecture: sensor → signal → data → interpretation → information for the user.
The digital layer works today. The laboratory layer is a research path – we show it as a sequence of stages rather than a finished result.
We are an interdisciplinary student research team exploring the intersection of chemistry, sensor technology, data science, artificial intelligence and digital technologies. We believe meaningful innovation often emerges where different disciplines meet.
Łukasiewicz–PORT (Polish Center for Technology Development, Wrocław) – cooperation on the laboratory component of the project.
Programme: Akademia STEM.
Our work combines scientific curiosity with practical experimentation. We develop hypotheses, examine possible solutions, construct demonstrators and refine the system architecture.
For questions, collaboration enquiries or further information.