Case Study: Engineering Group achieves optimal sequential decision-making with AnyLogic

Moving towards AI application for optimal sequential decision making

Engineering Group wanted to move beyond manual short-interval planning in a highly complex production line, where sequencing transformer-core movements could cause bottlenecks, line blockages, and costly crane-based resets. Working with AnyLogic, the company explored a digital twin approach using AnyLogic simulation software to replicate the shop floor, model pallets, conveyors, and production logic, and test better movement sequences.

AnyLogic implemented a simulation environment and then extended it toward reinforcement learning, allowing Engineering Group to evaluate “what-if” production plans and train policies for better decision-making. The result was a clearer way to investigate and rearrange sequences, anticipate issues before they happened, and reduce operating costs; in a simplified demo, the learned policy could move a core to its target position from any initial layout.

Engineering Group

Ramin Mirzazadeh

Engineering Group