MUMBAI, India, Sept. 14 -- Intellectual Property India has published a patent application (202611088393 A) filed by Uttaranchal University on July 20, 2026, for Ai Enable Real Time Food Contamination Detection System Using Multi Sensor Fusion.
Inventors include Sutakshi; Sweta Thakur; Rahul Mahala; Sonal Sharma; and Monisha Awasthi.
The application for the patent was published on September 11, 2026, under issue no. 37/2026.
Abstract: An AI-enabled real-time food contamination detection system and associated monitoring process utilize a multi-sensor data fusion network and cloud-based anomaly detection models to achieve automated, early-stage identification of insect infestations, microbial growth, and biochemical spoilage within storage infrastructures. The hardware system features a specialized smart sensory data collector device consisting of an edge microcontroller coupled to an array of environmental, physical, and chemical sensors, including an acoustic sensor to capture micro-vibrations from pests, a vibration sensor to detect crawling movements, a temperature probe to monitor heat spikes, a moisture probe to evaluate crop dampness, a humidity sensor to measure ambient air characteristics, and a gas sensor to track carbon dioxide, methane, and ammonia levels. Powered by an off-grid solar panel and a rechargeable battery managed by an automated power utility script, the device uses the edge microcontroller to execute data validation, noise filtering, and temporal formatting by combining metrics with an onboard real-time clock (RTC) and a GPS module before saving data to a backup SD card. Validated data packets are securely transmitted via a local Wi-Fi module to a cloud server that executes feature extraction algorithms to generate a unified feature vector representing the state of the food container. A cloud-hosted pre-trained machine learning model processes the feature vector to identify multi-parameter pattern deviations, computing a contamination severity index that eliminates false positives common to single-sensor frameworks.
Disclaimer: Curated by HT Syndication.
