engleski

Towards a knowledge-based Cognitive System for industrial application: Case of Personalized Products

The modern era brought new production paradigms and personalized production is among the most challenging ones. The integration of the customer's specific requirements into the product design is a challenge which can be addressed with the product configurator. In the mass production, the product configurators usually used the constraint-based approach which becomes insufficient for the B2C (business-to-consumer) relationship of the personalized production. Today, complex products result with complex product configurators which are not user-friendly. Nevertheless, the user is usually not interested in detailed technical specifications of the product, but rather in its performance. If the user could only define his/her needs in a term of product's performance, an intelligent backend algorithm should configure the product with desired performance. In this research, a step forward is made by designing a system, rather than designing an algorithm, which can take into account user preferences and decide which product configuration is the best compromise. This paper presents a prototype of the knowledge-based cognitive system for solving special case of product configuration problem. It is the performance-based B2C product configurator for the personalized production era. The proof of concept is the case of solving the product configuration instance with criteria which require humanoid decision-making.

Cognitive system ; Personalized production ; Knowledge system ; Multi-criteria decision-making ; HUMANT algorithm

nije evidentirano