MUMBAI, India, Aug. 10 -- Intellectual Property India has published a patent application (202641085455 A) filed by Sri Manakula Vinayagar Engineering College on July 13, 2026, for Gripsense-An Ai Driven Hybrid Robotic Arm Gripper.
Inventors include Dr. S. Anbumalar; Dr. D. Raja; Dr. D. Sivaraj; and K. Thangaraj.
The application for the patent was published on July 31, 2026, under issue no. 31/2026.
Abstract: The titled invention discloses an Al-based adaptive hybrid robotic arm gripper for intelligent object handling and autonomous gripping applications. The system integrates a camera module (2), a PC with Al processing unit (3), an Al model (4), an ESP32 controller (7), a robotic arm (13), and a hybrid gripper (18) incorporating multiple gripping mechanisms within a single robotic platform. The camera module (2) captures real-time images of objects, which are processed using image preprocessing techniques and classified by the Al model (4) comprising a CNN-ResNet-18 deep learning model to determine the object's material and geometric characteristics. Based on the classification result, an Al-based decision layer automatically selects the most suitable gripping mechanism from a mechanical gripper servo motor (19) for irregular and non-metallic objects, a vacuum (linear-actuator servo motor) (20) with an air vacuum pump (22) for smooth and flat objects, and an electromagnet (21) for ferromagnetic objects. The grip selection command is transmitted through the Wi- Fi/Serial communication module (6) to the ESP32 controller (7), which coordinates the movement of the robotic arm (13) and actuates the selected gripping subsystem through the servo driver (9) and motor driver (10). The robotic arm autonomously performs object gripping, lifting, placement, and release operations without human intervention. By combining multiple gripping mechanisms with Al-based object recognition and automatic grip selection, the invention improves gripping accuracy, operational flexibility, handling efficiency, and reliability while reducing manual intervention, downtime, and object damage.
Disclaimer: Curated by HT Syndication.
