Dynamic pricing of product and delivery time in multi-variant production using an actor critic reinforcement learning

The profitability of manufacturers in multi-variant production is challenged by the combination of increasing customer requirements and volatile supply chains. A potential solution is dynamic pricing, where customers can select a delivery time and price based on their preferences, and demand can be balanced during peak times. This paper presents a dynamic pricing approach using an actor-critic reinforcement learning agent in combination with a production simulation model and applies it in the automation technology industry.& COPY; 2023 CIRP. Published by Elsevier Ltd. All rights reserved.

Automated summary

The profitability of manufacturers in multi-variant production is challenged by increasing customer requirements and volatile supply chains. Dynamic pricing, where customers can choose delivery time and price based on preferences, is proposed as a potential solution to balance demand during peak times. This paper presents a dynamic pricing approach using an actor-critic reinforcement learning agent in combination with a production simulation model applied in the automation technology industry.