AI technology in use for process optimisation in production & logistics

Artificial intelligence is now an established technology for optimally designing or controlling processes in production, logistics, supply chains or other areas. There are now various AI processes that are used or offered in ready-made solutions (e.g. via Pathmind, Bonsai, H2O.ai):

• Reinforcement Learning
• Classifier

Challenge: Selection and training of a solution

The particular challenge in the application of AI methods is usually the selection and training of a solution (e.g. in-house development). As a rule, this cannot and should not be done on real systems, since extreme events can occur, especially in the first learning phase. In order to exclude this risk, simulation methods are used which, in combination with the AI application, represent an optimal solution:

• A simulation model provides realistic (dynamic) data of a possible scenario (typical challenge when using AI procedures).
• The model serves as a training partner to show decisions and their effects to the AI in a very short time.
• Extensive and risk-free pre-testing of AI solutions before they are released.
• Comparison of AI solutions or solution procedures.
• Quantitative evaluation of AI procedures to other possibly more cost-effective alternatives (classical optimisation procedures).

SimPlan provides the necessary technologies for such an environment, up to the point that we develop appropriate models and building blocks for our customers so that they can find, evaluate and risk-free introduce their AI application without in-depth simulation knowledge.