Proxima Technology Company Profile

We are pleased to announce a significant update to the user interface (GUI) features of our proprietary optimal control AI, "Smart MPC," in its embedded product version, "E-Smart MPC." With this update, you can complete everything from initial setup to operation using only the GUI, meaning no programming is required (no-code), allowing for a smooth user experience! For more details, please watch the video in the "Related Links"! *Related Links Video: How to Use: Setup and Control Demonstration in just under 2 minutes.

We will showcase the control AI "Smart MPC" that realizes optimization of manufacturing site control. ■ Event Overview Date: January 21 (Wed) - 23 (Fri), 2026, 10:00-17:00 Venue: Tokyo Big Sight, South Hall 4 Exhibition: Factory Innovation Week 2026 ■ Features of Smart MPC Achieves AI control without specialized knowledge through Model Predictive Control × Machine Learning - Energy-saving effect: 5-15% reduction - Implementation period: as short as 3 days (data collection operation starts) - Temperature accuracy: ±0.2℃ (PID ratio ±0.5℃ → ±0.2℃) - Automatic adjustment: Adapts to aging degradation through online learning ■ Solving these challenges ✓ Time-consuming PID parameter adjustments ✓ Difficulty in complex multivariable control ✓ Control quality instability due to lack of skilled personnel ✓ Desire to balance energy savings and quality A team of 36 engineers, including 20 with PhDs, will solve manufacturing site challenges using advanced mathematical optimization techniques. We will present live demonstrations and case studies at the venue. Please visit our booth.

Industrial furnaces are a major source of CO₂ emissions, accounting for about 15% of domestic emissions. Energy conservation and decarbonization have become unavoidable challenges for the manufacturing industry in terms of both cost reduction and environmental management. **Issues** Controlling industrial furnaces is complex. It requires precise adjustments of multiple heat sources during the heating, holding, and cooling processes, where even slight delays or deviations can lead to overheating and temperature inconsistencies. Traditional PID control struggles to flexibly respond to these complex thermodynamic changes and disturbances, resulting in accumulated structural energy losses due to "overreaction" and "delayed response." **Solution** In this seminar, we will introduce "Smart MPC," a model predictive control that combines machine learning and optimization technology. By utilizing past operational data, it achieves high-precision predictive control without the need for specialized tuning. The embedded controller "E-Smart MPC" allows for easy implementation with direct mounting to control panels and GUI operation. **Implementation Results** CO₂ and fuel reduction (up to 25%), improved control accuracy (up to 80%), and reduced setup time (up to 60%). **Target Audience** Production engineers, equipment managers, and site managers who want to promote energy conservation and decarbonization in thermal control equipment such as industrial furnaces and dryers.