Inspiration

The idea for ChemSecure Tracker was inspired by the growing incidents of lab accidents, chemical mishandling, and compliance failures in industrial labs, universities, and warehouses. While visiting labs, we realized that chemical tracking is still mostly manual, relying on registers or Excel sheets. This creates risks of: 1) Human error in logging and monitoring. 2) Expired chemicals causing waste and hazards. 3) Non-compliance with safety regulations. 4) Environmental harm from improper disposal. We wanted to leverage Agentic AI to address this gap and create a smart, automated safety solution.

What it does

ChemSecure Tracker is an AI-powered safety and compliance system for managing hazardous chemicals in labs, industries, and warehouses. Key Functions 1) Smart Chemical Registry – Automatically scans barcodes/QR codes and uses AI to classify and log chemicals.

2) Real-Time Risk Monitoring – Tracks expiry dates, unsafe storage conditions, and potential hazards, sending instant alerts.

3) Compliance Assistant – Provides AI-driven recommendations aligned with OSHA and Indian safety standards.

4) Predictive Insights – Uses machine learning to forecast shortages, wastage, and high-risk events.

5) Sustainability Angle – Integrates with energy forecasting to optimize plant operations and reduce risks.

How we built it

We structured the project in stages:

  1. Backend Development

*Python with Flask/FastAPI for core APIs. *Database schema for chemical registry and logs.

  1. AI/ML Integration

*LLM-based safety assistant for real-time guidance. *Forecasting models for predicting shortages and hazards.

  1. Computer Vision

*Barcode/QR scanning system. *AI image recognition for chemical labels.

  1. Frontend

*Web app using React. *Mobile app (Flutter) for portable scanning and alerts.

  1. Cloud & Deployment

*IBM Hybrid Cloud for scalability. *Secure database integration for compliance logging.

Challenges we ran into

Every project has hurdles, and ours was no different:

1) Data availability: Getting diverse chemical datasets for training vision models was tough.

2) Real-time monitoring: Balancing accuracy with speed for instant alerts.

3) Compliance mapping: Aligning AI recommendations with both global (OSHA) and local Indian standards.

4) Scalability: Designing the system to work across labs, universities, and industries with different infrastructures.

Accomplishments that we're proud of

1) Built a working AI-driven prototype that integrates chemical registry, computer vision, and compliance assistant in a single system.

2) Successfully implemented barcode/QR scanning + AI classification to detect and log chemicals.

3) Designed an LLM-powered safety assistant that provides real-time recommendations for safe storage and handling.

4) Developed forecasting models that predict shortages, wastage, and high-risk situations.

5) Created a scalable cloud-ready architecture that can be deployed across labs, industries, and universities.

What we learned

The practical challenges of applying Agentic AI to real-world industrial problems.

How to design intelligent agents that combine computer vision + machine learning + LLMs.

The importance of safety compliance standards (OSHA & Indian regulations) in building trustworthy AI solutions.

Skills in cross-platform development (React web app + Flutter mobile app).

Teamwork and time management—balancing research, coding, and presentation within deadlines.

What's next for ChemSecure Tracker

1)Collect more data to make AI more accurate.

2)Add IoT sensors (temperature, humidity, leakage detectors) for real-time monitoring.

3)Build a voice assistant (English + Hindi) so lab staff can use it easily.

4)Update the system for global compliance standards (EU, ISO).

4)Run pilot projects in colleges, pharma labs, and government safety boards.

5)Turn ChemSecure Tracker into a full product (SaaS) for industries and institutions.

Share this project:

Updates

Submission history