In production engineering, some of the most important decisions are made with incomplete information. For Jean-Christophe Barbier, one question kept