Lessons learned from an intelligent tutoring system for computer numerical control programming (CNC Tutor)

Abstract

Advances in CAD/CAM software and CNC machining have made the transition of design and machining seamless. Once a part is designed in a CAD format and a user specifies the machine tool needed for each machining pass, the CAD/CAM software can generate the G-code and the G-code can be fed into the CNC machine directly without any delay. There is no need to write G-code for each machining job. However, understanding G-code is still valuable, especially when a machining job does not run smoothly. Intelligent tutoring systems (ITS) have been shown to be successful in helping students to learn about math and physics subjects. However, relatively few ITS have been used to teach engineering subjects. The objectives of the paper are to (1) create an intelligent tutoring system to teach basic understanding of G-code, and (2) evaluate the learning gains from the system, and (3) summarize lessons learned from the implementation. The system has been evaluated by 91 undergraduate students. Results suggest that the CNC Tutor design is instructionally effective and that students' subjective impressions of the system are positive. It appears that we may continue to develop similar types of Intelligent Tutoring Systems for other engineering subjects. It also appears that CNC Tutor's explanations and feedback are a good fit for active, visual learners. Possible enhancements include the addition of more video and/or simulations to help learners to visualize abstract concepts.