MUMBAI, India, Oct. 12 -- Intellectual Property India has published a patent application (202641118418 A) filed by Koneru Lakshmaiah Education Foundation on October 03, 2026, for Self-Calibrating Adaptive Escalation Control For Distributed Ai Inference.
Inventor includes Sai Pradhun Gudipudi.
The application for the patent was published on October 09, 2026, under issue no. 41/2026.
Abstract: The present invention relates to a self-calibrating adaptive escalation control system and method for distributed artificial intelligence inference across device, edge and cloud computing tiers. The system processes inputs using a compressed neural network model and selectively escalates inputs to stronger remote models based on an adaptive uncertainty threshold. A probe scheduler verifies a randomized portion of retained inputs, while an inverse-propensity weighting unit generates an error estimate. A conformal update unit dynamically adapts the threshold toward a prescribed reliability target, and a budget governor regulates escalation expenditure. Drift attribution, fail-safe control and fleet-level pooling further enable adaptive, resource-efficient inference under changing deployment conditions.
Disclaimer: Curated by HT Syndication.
