A methodology for analysing human reasoning when performing a supervision task has been proposed in WAHRPS project (Worlds for assessing human reasoning in process supervision), conducted by INSERM 455 [1] . The main issue is the modelling of different human reasoning styles implemented as Reference Artificial Reasoners (RAR), and the study of decision sequences in front of several dynamic processes. In this paper, an adaptive tool able to generate different types of dynamical systems with

more »

... asing difficulty degrees is presented. The design and implementation of a reference artificial reasoner (RAR) based on sequential control, and the implementation of a RAR based on qualitative reasoning is also reported. Finally, a discussion about the comparison between human and artificial decision sequences, in front of the same dynamical process, is included.