LabSentinel AI

Displays simulated laboratory conditions, sensor readings, and recent safety assessments.
Shows hazard scenarios, sensor inputs, risk evaluation, and safety recommendations.

Inspiration

Laboratories can face hazards such as gas leakage, smoke, fire, overheating, and chemical incidents. Traditional systems mainly depend on individual sensors and fixed thresholds. We wanted to create a system that can understand the complete situation using multiple forms of data.

What it does

LabSentinel AI combines camera input, sensor data, and voice/text input to detect and understand laboratory hazards. It assesses the risk, identifies people who may be affected, and provides suitable safety instructions. Generative AI helps create clear alerts and incident reports, while Agentic AI supports decision-making and response actions.

How we built it

We designed a multimodal system using environmental sensors, camera-based monitoring, and voice/text input. Sensor data is collected through a microcontroller and sent to the software platform. AI models process visual, sensor, and language inputs, while an AI agent combines the results to assess the situation and recommend actions. A web dashboard is used to display alerts and safety information.

Challenges we ran into

The main challenge was combining different types of data into one system. Sensor readings, camera information, and voice inputs have different formats and reliability. Designing an AI workflow that can correctly combine these inputs and provide useful safety actions was another challenge.

Accomplishments that we're proud of

We developed a concept that goes beyond traditional threshold-based laboratory monitoring. LabSentinel AI brings together multimodal AI, Generative AI, and Agentic AI to create a smarter and more proactive safety system.

What we learned

We learned how multimodal AI can combine different data sources to understand real-world situations. We also explored how Generative AI can create useful responses and how Agentic AI can support multi-step decision-making.

What's next for LabSentinel AI

We plan to improve hazard detection, add more sensors, improve real-time camera analysis, and enable automated alerts. We also aim to add multilingual voice support and improve the AI agent's ability to respond to complex laboratory emergencies.