Bayesian optimal experimental design for circular business models

ABSTRACT:

Implementing circular business models (CBMs) like Product-as-a-Service entails high uncertainty, necessitating costly and prolonged business experimentation. To efficiently mitigate this uncertainty, Bayesian Optimal Experimental Design is applied to the CBM context, selecting conditions that maximize the Expected Information Gain for unknown CBM parameters. Applied to an air conditioner subscription case study, the method successfully identified optimal conditions from 124 candidates. This approach facilitates CBM implementation by efficiently minimizing uncertainty under limited resources.