MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202611054216 A) filed by Indian Institute Of Technology Banaras Hindu University, Varanasi on April 28, 2026, for System And Method For Controlling Robots In Stochastic Environments.
Inventors include Samuel, Cherian; and Arya, Aryan.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: A system (100) and method (200) for controlling a plurality of robotic units (102) in stochastic environments using decentralized reinforcement learning and shared spatial intelligence. The system (100) integrates the plurality of robotic units (102), each including modular chassis (106) and a sensor module (108) that includes vision sensor, with a control module (104) configured to execute autonomous collective coordination. The system (100) determines proximity thresholds and alignment vectors via a reinforcement learning module (110). The system (100) actuates drive systems to dynamically adapt a collective load-bearing footprint for transporting payloads by seamlessly interlocking the modular chassis (106) into a continuous flat-plate surface. The system (100) generates three-dimensional Cartesian coordinates from real-time trigonometric projections of sensor data to maintain an asynchronous shared spatial intelligence map within a database (116). A voxel mapping module (112) and gesture interface module (114) utilizes shared map and vision data to execute preemptive collision avoidance.
Disclaimer: Curated by HT Syndication.
