Smart algorithms in production planning: Flexible planning for the shop floor

The Challenge: Flexible Algorithms for Optimized Production Planning

The multitude of complex constraints combined with practical problem sizes presents a significant challenge. To ensure optimal planning for each client, the system must automatically adapt to their specific situation. Since every company pursues different goals, these must be aligned to create a coherent overall plan. In addition, manual inputs from users must be incorporated into the planning.

The Solution: Modern Algorithms, Automatic Configuration, and Integration of Predictive Models

The project resulted in a holistic solution in which modern scientific methods, including AI and reinforcement learning approaches, were specifically expanded to make them usable for practical problems. The complex configuration and customization to each manufacturing company is significantly simplified through automated optimization of planning parameters (hyperparameter optimization). Predictive models can also be integrated directly into the planning process—models that are continuously trained using real-time data generated during production.

The Benefit: Better Solutions for Customers

The developed solution takes planning capabilities to the next level with the following features:

• Consideration of a wide range of constraints
• Optimal resolution of practically relevant problem sizes
• Automatic adaptation to the specific planning situation
• Integration of predictive models into the planning process
• Automatic configuration through automated optimization of planning parameters (hyperparameter optimization)

 Thanks to defined interfaces, additional features can be easily integrated.