EcoSenseAI reimagines environmental monitoring as a software-first robotic system, fully operational in simulation to align with AI-robotics integration. At its core, the platform simulates robotic sensors (e.g., via Raspberry Pi scripts with mock data for CO2 levels from 300-800ppm and temperatures 20-30°C) to mimic real-world IoT deployments without physical hardware. Data is ingested through AWS IoT Core (adaptable to cloud simulators), processed by a TensorFlow LSTM model that trains on historical datasets to predict CO2 spikes—alerting if levels exceed 600ppm for proactive interventions. The backend, powered by FastAPI, handles real-time ingestion, AI predictions, JWT authentication, and PDF report generation using ReportLab, with all data stored in MongoDB for scalable querying. A React frontend provides an interactive dashboard with Chart.js visualizations for trends, Material-UI alerts for spikes, and WebSocket integration for live updates, ensuring operators can monitor simulated robotic behaviors in a web-accessible app. This simulation-first approach addresses key robotics challenges: it enables low-cost training and evaluation pipelines (Track 2), where AI autonomously controls monitoring tasks, reacts to environmental changes, and generates analytics for industries like healthcare facilities or manufacturing warehouses. For instance, in a virtual digital twin of a warehouse, the system could simulate robotic sensors detecting air quality issues, triggering automated alerts or adjustments.
Category tags: