MUMBAI, India, July 24 -- Intellectual Property India has published a patent application (202631079841 A) filed by Dr. Rahmuddin Miyan; R. Roopa; Ms. Shalu Kumari; Dr. M. Manickam; Mr. R. Chinnadurai; Dr. Y. Sameena; Dhanasekar S; Dr. Palanivendhan; and K T Evangelin Monica on June 29, 2026, for Deep Learning-Driven Adaptive Human-Robot Collaboration System For Real-Time Safety And Productivity Optimization In Industrial Environments.
Inventors include Dr. Rahmuddin Miyan; R. Roopa; Ms. Shalu Kumari; Dr. M. Manickam; Mr. R. Chinnadurai; Dr. Y. Sameena; Dhanasekar S; Dr. Palanivendhan; and K T Evangelin Monica.
The application for the patent was published on July 17, 2026, under issue no. 29/2026.
Abstract: The present invention discloses a Deep Learning-Dril,en Adaptive Human-Robot Collaboration System fbr Real-Time Safety and Productivity Optimization in Industrial Environments. The system is designed to enable intelligent. safe, and efficient collaboration betr.t'een human operators and industrial robots within shared workspaces. The invention integrates a multi-modal sensor framework comprising RGB cameras. depth sensors, LiDAR sensors, wearable devices, force-torque sensors. and environmental sensors to continuously acquire real-time operational data. A sensor fusion rnodulc synchronizes and aggregates heterogeneous data streams, which are processed by a deep learning analysis engine to perform human detection, activity recognition, motion trajectory prediction. intention estimation, fatigue assessment, and hazard identification. Based on the analyzed information. a predictive safety module generates a Dynamic Context-Aware Safety Index (DCSI) representing the real time risk level associated with human-robot interaction. An adaptive decision-making engine utilizes the generated safety index and reinforcement learning techniques to dynamically adjust robot operational parameters including speed, trajectory, workspace boundaries, and task allocation. The system further establishes Adaptive Virtual Safety Zones (AVSZ) that continuously modify safety boundaries according to predicted worker behavior and environmental conditions. A productivity optimization module minimizes idle time, balances workloads, and enhances manufacturing throughput through intelligent task scheduling and resource allocation. The proposed invention continuously leams from historical and real-time operational data, enabling self-improving collaboration strategies without manual reconfiguration. The invention provides a unified framework for predictive safety management and productivity enhancement in smart manufacturing, warehouse automation. logi.tics, and Industry 4.0 applications, thereby improving worker safety, operational efficiency, and overall system performance.
Disclaimer: Curated by HT Syndication.
