Оцінка можливостей EDGE-обчислень та AI-інференсу в промислових контролерах нового покоління (Beckhoff CX ТА MITSUBISHI IQ-R з AI-модулями)

Abstract

This paper evaluates edge computing and AI inference capabilities in nextgeneration industrial controllers, focusing on Beckhoff CX series (PC-based) and Mitsubishi MELSEC iQ-R with AI modules. Key features such as real-time inference engines, hardware acceleration (GPU/NPU), TwinCAT Machine Learning for Beckhoff, and AI visual inspection/integration for Mitsubishi are analyzed. Results show significant advantages in predictive maintenance, vision tasks, and low-latency decisions at the edge, supporting Industry 4.0 requirements in 2026