MUMBAI, India, Sept. 14 -- Intellectual Property India has published a patent application (202641095100 A) filed by The Principal, Mepco Schlenk Engineering College on August 06, 2026, for A System And Method For Real Time Emotion Detection And Adaptive Behavioral Nudge Delivery Based On Artificial Intelligence.
Inventors include Dr. K. Rajamani; and S. Lirish Lenny Paul Fdo.
The application for the patent was published on September 11, 2026, under issue no. 37/2026.
Abstract: ABSTRACT OF THE INVENTION A system and method for real time emotion detection and adaptive behavioral nudge delivery based on artificial intelligence. The present invention provides for an Emotion Aware Adaptive Nudge Engine (EANE) that detects emotions in real time, and delivers adaptive nudges based on those emotions. The system fills a major gap in current behavioral intervention technologies by bringing nudges at the right time when the user's emotional and physiological state are optimal for the desired behavior and intervention to maximize the effectiveness of the intervention and avoid negative psychological reactions. The system is divided into five modules which are inter functionally integrated. Modality specific deep neural network encoders are used to continuously process voice prosody, facial micro expressions, keystroke dynamics and physiological signals from the wearer (including galvanic skin response and heart rate variability). The Affective State Fusion and Classification Layer is a multi-head attention mechanism that is designed to combine the vectors of the features of the modalities and to classify the state of the user in terms of one of six classes: Calm, Stressed, Anxious, Motivated, Fatigued, or Neutral. The Nude Eligibility and Suppression Engine compare the classified state to a machine learned receptivity threshold, and actively suppresses the delivery of the nudge when the user is in an "unreceptive" state, passing the window of delivery to an adaptively scheduled future time. The LLM Based Adaptive Content Generator dynamically generates nudge messages using a structured prompt engineering schema, which are calibrated to the detected emotional context with respect to the tone, linguistic register and format of the messages. The Reinforcement Learning Feedback Engine transforms the nudge delivery problem into a Markov Decision Process (MDP) and uses proximal policy optimization to learn the policy of delivery of the timing and choice of nudge content from longitudinal user response data. The key novel contributions of the invention are: the real time detection of emotions and nudge suppression logic; the dynamic adaptation of nudge content to the emotional state classification; and the use of reinforcement learning for the paired (emotional context, nudge outcome) training paradigm. The system is designed to process the data at the edge of the device while maintaining user privacy. The applications include healthcare adherence, mental wellness support, financial behavioral coaching, productivity enhancement platforms, among others.K
Disclaimer: Curated by HT Syndication.
