Artworks transportation processes are highly challenging operations between multiple partners that are specified over service level agreements. They are executed within dynamic and uncontrollable environments that might leave undesired damages on the artworks if not handled efficiently. In this paper we take advantage of the emerging concept of IoT, and we propose PACT-ART (stands for "Processus Adaptatif au Contexte pour le Transport d'oeuvres d’ART" in French.), realized as an architecture that employs advanced computing techniques like data mining and business process intelligence to predict a future state of the process and point out any possible violation from it. Then, by recommending dynamic adaptation actions, the foreseen misbehavior could be eliminated or mitigated proactively.