Michael Pennell

Title: Michael Pennell and Hybrid Reinforcement Learning for Water Distribution Networks.

Introduction

Michael Pennell is an inventor based in San Clemente, California, and is associated with Autodesk, Inc. His patent work applies reinforcement learning to the control of water distribution networks, addressing operational goals such as maintaining tank levels and reducing energy costs.

Latest Patents

Pennell is credited with 1 patent: “Hybrid Reinforcement Learning (RL) to Control a Water Distribution Network.” The invention describes a method and system that use historical operating states, residential water demand, tank levels, and energy tariffs to guide network control. A query-based model identifies control points, while a reinforcement-learning agent is trained using prior states and rewards that account for tank-level constraints, energy costs, and pump toggles. The agent selects time slots and generates hybrid setpoints, including pump-speed adjustments, to help maintain tank levels, minimize energy costs, and comply with toggle limits.

Career Highlights

Pennell’s listed company affiliation is Autodesk, Inc. His patent reflects an application of machine learning to infrastructure operations, combining historical data and adaptive control to manage water distribution.

Collaborations

Yuan Zhou and Shu Wang are identified as Pennell’s coworkers. Their names are provided as professional associations, though the available information does not specify their roles in the patent.

Conclusion

Michael Pennell’s patent presents a reinforcement-learning approach to balancing water-system reliability, energy efficiency, and equipment operation constraints.