Traditionally, in industrial production systems, the reconfiguration of machinery relies on the expertise of human operators. This expertise is accumulated over time as the worker gains experience. The actions and intentions of experienced operators can be recorded with motion capture technologies such as body and eye-gaze trackers. This data can potentially support inexperienced or less proficient workers in acquiring knowledge to perform similar tasks by simulating the actions of the experts. This work presents a method to capture and analyze the motions of an expert operator while reconfiguring machine tools of the forming machine. The findings reveal that actions associated with reconfiguring machine tools for forming press machine can be recognized and segmented by applying time-domain analysis on the motion data. The results show potential of developing a knowledge-based assistance system that encapsulates the expertise of operators and supports novice operators to replicate learned actions effectively in order to enhance productivity on the shop floor.