Keio University

ChatMPC: Control System Improvement via User-Machine Conversation

Published: June 02, 2026
Public Relations Office

Keio University

Associate Professor Masaki Inoue and Assistant Professor Yuya Miyaoka (fixed-term, PhD 1st year) of the Keio University Department of Applied Physics and Physico-Informatics, in collaboration with Professor Jose M. Maestre, have developed a technology called ChatMPC.

ChatMPC enables real-time improvement of control systems through simple, casual, and conversational interaction with users. By integrating an AI model that interprets verbal interactions with a control algorithm that ensures system safety, the system can remain reliable, even under ambiguity in user instructions or AI fluctuations.

These research results were published as a rapid release in the IEEE journal Transactions on Control Systems Technology on 30 May, 2026 (Japan Time).

For further information, please refer to the following PDF file.

Press Release (PDF)