Creating and Implementing a Custom Chatbot in Engineering Education

This paper investigates the development and use of a chatbot in an engineering curriculum. The chatbot helps students find course materials, answer general inquiries, schedule meetings with professors and teaching assistants, and much more. Students require assistance during their time at university. College life is stressful, and tasks such as keeping track of deadlines, scheduling meetings, and finding resources become daunting as the semester progresses. The constant email exchanges about general course information could also become tedious for the course instructors. The authors developed a chatbot using the Microsoft Power Virtual Agents App, which requires minimal coding, as the main framework consists of configuring and connecting nodes. The development stages require some considerations, such as setting up a hierarchical system to store information about various courses and creating a shared department email address to send and receive meeting requests. These considerations are addressed further in the paper. The chatbot will handle all repetitive tasks, thus freeing the instructors time for answering more challenging questions. It also promotes self-learning and allows students to ask questions beyond office hours and get responses. The chatbot developed is set up to source information from existing databases of deadlines, file directories, questions, and answers. Hence, it doesn't require reprogramming for a new course. The instructor needs to tabulate all necessary information in a predetermined location. Future variations of the chatbot will help students with coursework by providing them with practice problems when prompted. It will assist them further by giving instant feedback on their questions. Thus, students will get to practice and prepare for a course as they see fit. Since this process will be automated and randomized, the chatbot will always have sample problems. The chatbot will record any question that it is unable to answer. The instructors can further develop the chatbot's repository to answer those questions. The chatbot's impact on the student's university experience is measured in a class by conducting class surveys among the students. The authors have planned a pilot study of the chatbot and its implementation for a course in Spring 2023. Results will be reported in the final paper.