Patent Portfolio

Agentic AI Control System for Autonomous Laser Operations in Semiconductor Lithography

Inventors: Jacob Peter Strock, Yuji Minegishi
Submission Date: February 4, 2025
Patent Office: United States (US)
Application Number: 20250291338

Description
A containerized, agentic AI system for managing semiconductor laser infrastructure. The platform combines LLM-based reasoning, tool-using agents, and real-time control pipelines to enable autonomous query handling and direct operational control of laser systems.

Abstract
A laser management server for a laser device includes a sending and receiving processor receiving a query from outside; a query input processor receiving and decomposing the query, and generating a first query item requiring external information and a second query item without requiring external information; an agent action processor receiving the first query item, acquiring the required external information from unstructured data and structured data, and generating a first agent response; an equipment agent processor receiving the second query item, receiving the first agent response, and generating a query item prompt related to the second query item; a large language model processor receiving the query item prompt, and generating a query item response; and an equipment control processor receiving a first control signal, and transmitting the first control signal to the laser device. The equipment agent processor receives the query item response and generates the first control signal.

Deep Learning and Time Series Forecasting Framework for Predictive Hardware Replacement in Lithography Laser Systems

Inventors: Jacob Peter Strock, et al.
Filing Date: November 6, 2025
Patent Office: United States (US)
Application Number: US-20250342342-A1

Description
A deep learning-based time series forecasting system designed to model and predict laser performance under varying maintenance and hardware replacement scenarios. The framework integrates multi-sensor temporal data streams (e.g., pressure, voltage, usage cycles) to simulate future system states.

Impact

  • Predictive maintenance planning driven by temporal modeling
  • Lifecycle optimization of high-cost hardware components
  • Reduced downtime through forward-looking system performance estimation
Real-Time Digital Twin Platform with Deep Learning-Driven Forecasting and Time-Aligned Visualization

Inventors: Jacob Peter Strock, et al.
Filing Date: October 30, 2025
Patent Office: United States (US)
Application Number: US-20250336118-A1

Description
A digital twin platform that integrates real-time telemetry with deep learning-based forecasts to create a time-aligned, continuously updating simulation environment. The system enables synchronized visualization of actual and predicted system behavior.

Impact

  • Continuous monitoring of system state vs. predicted trajectories
  • Scenario simulation using learned temporal dynamics
  • Unified interface for operational decision-making and optimization
Neural Network-Driven Control Optimization for Dynamic Parameter Tuning in Semiconductor Laser Systems

Inventors: Jacob Peter Strock, et al.
Filing Date: October 30, 2025
Patent Office: United States (US)
Application Number: US-20250335767-A1

Description
A deep learning-based control optimization system that models the impact of operational parameter changes (e.g., gas pressure, voltage) on system performance. The model enables real-time inference for safe and optimal parameter adjustments.

Impact

  • Data-driven parameter tuning with predictive guarantees
  • Stabilization of complex physical systems under changing conditions
  • Improved operational efficiency through learned system dynamics