Intellectual Property India Publishes Patent Application for An Intelligent Multimodal Edge-Computing System And Method For Real-Time Non-Destructive Shelf-Life Prediction Of Agricultural Produce Using Computer Vision, Chemical Sensing, And Sensor Fusion

MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202641094830 A) filed by Sns College Of Technology on August 05, 2026, for An Intelligent Multimodal Edge-Computing System And Method For Real-Time Non-Destructive Shelf-Life Prediction Of Agricultural Produce Using Computer Vision, Chemical Sensing, And Sensor Fusion.

Inventors include Dr. J. Geetha Ramani; Dr. K. S. Vairavel; Dr. D. Pradhabhan; Dr. T. C. Kumari Sugitha; Mr. G. Durgai Raja; Mr. K. Arikaran; and Mr. P. Dhinagar.

The application for the patent was published on August 14, 2026, under issue no. 33/2026.

Abstract: The present invention relates to an intelligent multimodal edge-computing system and method for real-time, non-destructive prediction of the remaining shelf life of agricultural produce. The proposed system comprises an Edge Processing Unit (1), a USB High-Definition Camera (2), an Arduino Uno Microcontroller (3), a pH Sensor Module (4), an MQ-Series Ethylene/VOC Gas Sensor (5), a Temperature and Humidity Sensor (6), a Produce Placement Chamber (7), a Sensor Fusion Engine (8), a Deep Learning Inference Module (9), a Shelf-Life Prediction Engine (10), an LCD Display Module (11), a Graphical Dashboard (12), a Communication Interface (13), and a Power Supply Module (14). The Edge Processing Unit (1) synchronously acquires image data from the USB High-Definition Camera (2) and biochemical and environmental measurements from the Arduino Uno Microcontroller (3) through the Communication Interface (13) to generate a temporally synchronized multimodal dataset. The Deep Learning Inference Module (9) extracts visual features associated with ripening and spoilage, while the Sensor Fusion Engine (8) integrates optical, biochemical, and environmental information to generate a Unified Freshness Score. The Shelf-Life Prediction Engine (10) estimates the remaining shelf life by dynamically compensating for temperature and humidity variations. Prediction results are displayed through the Graphical Dashboard (12), while real-time sensor measurements are presented on the LCD Display Module (11). The invention provides an accurate, portable, scalable, and low-cost solution for food quality assessment, inventory management, and reduction of post-harvest losses in agricultural supply chains.

Disclaimer: Curated by HT Syndication.