Back to Nilesh's profile
8views

GreenNexa

GreenXena is an AI-powered IoT platform that transforms real-time facility data into intelligent insights for smarter, more efficient, and sustainable resource

Nilesh Panigrahy
  • Problem, approach and result explained
  • Expert reviewed· 6 reviews
GreenNexa

Overview

The problem

Indian cities and institutions face challenges in managing energy consumption, water usage, air quality, waste management, and other facility resources. Currently, these resources are often monitored separately, making it difficult for administrators to identify inefficiencies, detect problems early, and take timely action.

Our project addresses this problem by developing a unified intelligent dashboard that collects real-time data from multiple sensors and provides actionable insights to help municipalities and facility managers improve resource efficiency and sustainability.

A good solution should reduce resource wastage, improve operational efficiency, and support data-driven decision-making.

The approach

We developed GreenXena using a combination of IoT, full-stack development, and AI/ML technologies.

Data Sources:

  • ESP32-based IoT sensor nodes.
  • Energy and solar power monitoring sensors.
  • Water-level sensors.
  • Air-quality sensors for environmental monitoring.
  • Temperature and humidity sensors.
  • Smart dustbin monitoring sensors.
  • Historical and synthetic datasets for testing and analysis.

Technology and Features:

  • React.js for the frontend dashboard.
  • Node.js and Express.js for backend APIs.
  • Database integration for storing and retrieving sensor data.
  • Python and PyTorch for AI/ML processing.
  • Real-time monitoring and visualization.
  • Anomaly detection to identify unusual resource consumption.
  • Time-series forecasting to predict future trends.
  • AI-based recommendations for improving resource efficiency.
  • Priority Engine for identifying critical issues.
  • Alerts, sustainability scoring, and facility-wise monitoring.

Model Selection:

We use an Autoencoder-based approach for anomaly detection and LSTM-based time-series forecasting.

We selected these approaches because they can learn patterns from sensor data, identify unusual behaviour, and capture changes over time. Compared with simple threshold-based monitoring and basic statistical methods, they provide more intelligent analysis and predictive capabilities.

Our architecture connects IoT devices, backend APIs, database storage, and AI/ML services into one integrated system.

The result

GreenXena delivers a unified dashboard for monitoring and managing facility resources.

Our prototype demonstrates:

  • Real-time sensor data collection and visualization.
  • Centralized monitoring of energy, water, air quality, and waste.
  • AI/ML-based anomaly detection and forecasting.
  • Intelligent recommendations for resource optimization.
  • Priority-based alerts for critical issues.
  • Sustainability insights to support better decision-making.
  • Integration of real sensor data and synthetic data for testing.

We successfully developed and tested the major components of our prototype, achieving 35/35 passing tests.

The solution demonstrates how IoT and AI/ML can work together to make facility management smarter, more efficient, and more sustainable. It is designed to be scalable for future deployment across multiple facilities and municipalities.

Reflection

During the development of GreenXena, we learned how to integrate IoT hardware, backend services, databases, and AI/ML into a unified platform.

If we could improve our project, we would focus on making the system more scalable, improving sensor accuracy, and collecting more real-world data to improve the reliability of our AI/ML predictions.

Some parts, such as large-scale multi-facility deployment and long-term forecasting accuracy, still need further testing and improvement. We also learned that real-world sensor data can be inconsistent, so proper data validation, error handling, and sensor calibration are essential.

One of our biggest learnings was that developing a smart solution is not just about building individual features; it is about connecting every component reliably and making the output useful for real-world users.

In the future, we aim to improve the prediction models, strengthen system security, optimize performance, and expand GreenXena for deployment across multiple facilities and municipalities.

Submitted to Sustainable Facility & Estate Intelligence

Your idea in one line

AI-powered Sustainable Facility and Estate Intelligence Dashboard that monitors energy, water, air quality, waste, and traffic to predict inefficiencies, detect anomalies, and recommend data-driven actions for smarter, greener infrastructure.

What will be working by the finale?

Our project is a sensor-based sustainability monitoring system that collects real-time data from sensors, visualizes energy, water, air quality, and other environmental parameters, detects anomalies using deep learning and forecasts potential issues, and provides actionable recommendations by using priority engine for efficient resource management.

Tech stack and hardware you plan to use

In our project we use for frontend :-react js and css For Backend :Express js For database :MongoDB for API :Rest API(Fast api) For Deep Learning : Pytorch(AUTO ENCODER, LSTM) for sensor : 2 Water flow sensor 2 solenoid valves 4 ESP 32 micro controller 2 channel relay 2 peice 12 v DC pump Water level sensor Solar charge controller ACS712 current sensor -2 2 voltage sensor 3 lithium ion cell 2000mah 3S BMS 5 volt voltage regulator Mq135 ,mq2,mq9,DHT22,SGP30 ,BMP 280 Rain sensor Solar panel 20 watt Ultrasonic sensor 2 peice

What's working now?

GreenNexa is an IoT-based Sustainable Facility and Estate Intelligence Dashboard that integrates a physical hardware prototype with a real-time software platform. What's working: 1. IoT Hardware: We have developed a physical smart facility prototype with sensors to collect actual environmental and facility-related data. 2. Real-Time Data Monitoring: The system collects real sensor readings and processes them for monitoring through the dashboard. 3. Dashboard: A centralized dashboard visualizes key parameters, including energy consumption, water usage, air quality, waste management and occupancy. 4. Backend & Database: FastAPI-based REST APIs handle sensor data communication, while the database stores and manages the collected information. 5. AI/ML Integration: An LSTM Autoencoder is used for anomaly detection, and an LSTM-based forecasting approach estimates future energy consumption. What's partially working: - Further optimization of hardware-to-dashboard communication. - Additional validation and performance testing of AI models using larger real-world datasets. What's not completed yet: - Full-scale deployment in actual buildings. - Long-term field testing and production-level scaling.

A photo or screenshot
A photo or screenshot: GreenNexa

AI tools used

Claude

for research purpose

Gallery

5