Session: 09-08-01: Distance/Online Engineering Education, Models and Enabling Technologies

Paper Number: 112683

112683 - "Implementation of an Artificial Intelligence (AI) Instructional Support System in a Virtual Reality (VR) Thermal-Fluids Laboratory" 

Physical laboratory experiments have long been the cornerstone of higher education providing future engineers practical real-life experience invaluable to their careers. Due to the large influx of engineering students in combination with the shortage of staff availability post-COVID, demand for laboratory time has exceeded physical capabilities with maintenance of existing laboratories and investment in new equipment falling behind. Virtual reality (VR) labs have proven to retain many of the benefits of attending physical labs while also providing significant advantages only available in a VR environment. The potential for VR to supplement existing physical lab curriculum in higher education has been acknowledged but the pace of adoption has been slow.  

Previously, our group had developed a pilot VR lab that aimed at improving the quality of education through an immersive experience of existing fluid mechanics experiments. The VR lab contained six (6) unique thermal fluids lab experiments developed using the Unity game engine. We have since then optimized the performance and unified the appearance across all labs. The VR lab allowed users to interact with simulated lab setups where they could learn the experimental procedures, record additional data sets, and dive deeper into the content at their own pace. 

In this paper, we discussed the preliminary testing results of conducting one of the VR labs in the fluid mechanics laboratory class across two groups. Seventeen (17) students in the experimental group used the VR lab to conduct the experiment before writing the lab report, whereas the fourteen (14) students in the control group were not exposed to the VR lab. The overall feedback from the experimental group was favorable, with students showing positive interest in using VR lab simulations to supplement and expand their laboratory education. Lab report quality was also compared between the experimental group and the control group. We further discussed a new development in this VR lab: An Artificial Intelligence (AI) instructor. The AI instructor was developed in Unity as a non-player character (NPC) that monitored the state of the lab experiment and became interactable to voice and text based on an in-game proximity sensor. In order to provide a wider range of interactions and realism, the NPC was provided a connection to the Generative Pre-trained Transformer 3.5 (GPT-3.5) language model through publicly available API’s from OpenAI. The AI instructor became capable of providing instructions with pre-loaded lab context and identifying users’ mistakes in conducting the lab procedure to provide an experience similar to a real-life lab instructor. Survey results were collected from both groups to gather feedback on the intuitiveness, functionality, and potential resourcefulness of the updated VR lab in an educational setting. 

Based on the survey data from the initial test and our subsequent changes to the existing VR labs, the challenges limiting the potential of VR in education are addressable with creative solutions using already present tools within the Unity game engine. As iterations to the VR labs continue to refine the experience and reduce barriers for its usage, the roadmap is being laid for additional educators to make use of this rapidly developing technology to provide a more enriching experience for all students who benefit from learning in engaging settings of different forms. 

Presenting Author: Dennis Ayre California State Polytechnic University Pomona

Presenting Author Biography: Dennis Ayre is an undergraduate student at California State Polytechnic University – Pomona studying Mechanical Engineering with an emphasis on thermal-fluids. He is using his past experience in computer science to develop new software applications for both academic and industry use cases. His current research involves exploring the potential for new technologies in the educational space including the responsible use of AI. Dennis is a current resident of Encino, CA and enjoys mountain biking and off-roading in his spare time.

Authors:

Dennis Ayre California State Polytechnic University Pomona
Carolyn Dougherty California State Polytechnic University Pomona
Yitong Zhao California State Polytechnic University Pomona