Board 49: Work in Progress: Using Generative AI for Reducing Faculty Workload in Online Engineering Courses

Gerry A Pedraza; Sunay Palsole

Download Paper | Permalink

Conference: 2024 ASEE Annual Conference & Exposition
Location: Portland, Oregon
Publication Date: June 23, 2024
Start Date: June 23, 2024
End Date: July 12, 2024
Conference Session: Computers in Education Division (COED) Poster Session
Tagged Division: Computers in Education Division (COED)
Permanent URL: https://strategy.asee.org/47045

Request a correction

Paper Authors

biography

Gerry A Pedraza Texas A&M University

visit author page

Gerry is the Assistant Director of Learning Design at the Engineering Studio for Advanced Instruction and Learning at Texas A&M University. He is a proactive innovator dedicated to enhancing faculty workflows in collaboration with instructional designers. His primary goal is to streamline faculty transition to online teaching, fostering seamless interactions between educators and instructional staff.
Gerry's work is instrumental in saving valuable time for creative and impactful educational endeavors, ultimately benefiting both educators and students in the engineering field. His dedication extends to ensuring that faculty and instructional designers are equipped with various choices, enhancing their ability to tailor educational content to the needs of diverse learner communities.

visit author page

biography

Sunay Palsole Texas A&M University

visit author page

Dr. Palsole is Assistant Vice Chancellor for Remote Engineering Education at Texas A&M University, and has been involved in academic technology for over 20 years. He helped establish the Engineering Studio for Advanced Instruction & Learning (eSAIL),

visit author page

Download Paper | Permalink

Abstract

WIP. The demand for high-quality online engineering courses and credentials is surging, driven by the upskilling and reskilling needs of industry partners and engineers with 8-10 years of experience. Creating accessible, top-tier online courses require producing exceptional videos, transcripts, and content segmented appropriately for optimal student learning[1]. Beyond lecture preparation, faculty are often tasked with creating well designed slides and assessments to engage students and measure learning. Despite the support of instructional designers in many institutions, this multifaceted process presents a significant challenge for faculty, who are often concurrently engaged in research, service duties, and mentoring activities [2], [3]. To support instructional designers and faculty in this endeavor, we have leveraged the APIs of OpenAI tools to create Transcriptto, a Python program that contains clever algorithms that aid in the crucial steps in lecture preparation, allowing instructional designers and faculty to have a better starting point when starting the development of an online course. Transcriptto utilizes a straightforward yet robust workflow, incorporating openly available technologies such as Pymovie, FFmpeg, OpenAI’s Whisper, and ChatGPT. It transforms video lectures into polished text, supporting various input types, including audio files, and pre-existing scripts. The tool enhances clarity, removes redundancy, student interventions, semester-specific details, and provides a user-friendly configuration file for easy customization without requiring programming knowledge. The resulting content is optimized for information processing and fine-tuned via the OpenAI API, yielding not just a polished transcription that may be used to record a streamlined lecture more efficiently, but a multifaceted educational suite that generates outlines for slides, quizzes, assignments, and discussion questions. All learning objects are based on the content being taught. Preliminary test usage of Transcriptto indicates faculty are enthusiastic about options for enhancing existing lecture materials and creating an optimized experience for their online course. We estimate this process will contribute to cutting the time it takes to write a script for studio sessions, and faculty will also appreciate the efficient creation of high-quality additional learning objects to evaluate students and maintain engagement. To assess the efficacy of Transcriptto, we employ a modified schema after Merriam [4], conducting a survey and focus groups with faculty users who provided their content as input for the program to process. Our aim is to gather in-depth insights, experiences, strengths, weaknesses, and suggestions related to the tool, actively facilitating discussions, and analyzing qualitative data to draw conclusions about its efficacy. The tool’s adaptability, driven by ongoing feedback from faculty, instructional designers, and other experts in the field, ensures its alignment with real-world educational needs. Keywords: AI, Instructional Design, Online Transition, Online Education, e-learning, OpenAI, Python, Time Saving Strategies for Faculty, Lecture Capture, Repurposing Video Lectures.

Citation
Format

Pedraza, G. A., & Palsole, S. (2024, June), Board 49: Work in Progress: Using Generative AI for Reducing Faculty Workload in Online Engineering Courses Paper presented at 2024 ASEE Annual Conference & Exposition, Portland, Oregon. https://strategy.asee.org/47045

TY  - CPAPER
AB  - WIP. The demand for high-quality online engineering courses and credentials is surging, driven by the upskilling and reskilling needs of industry partners and engineers with 8-10 years of experience. Creating accessible, top-tier online courses require producing exceptional videos, transcripts, and content segmented appropriately for optimal student learning[1]. Beyond lecture preparation, faculty are often tasked with creating well designed slides and assessments to engage students and measure learning. Despite the support of instructional designers in many institutions, this multifaceted process presents a significant challenge for faculty, who are often concurrently engaged in research, service duties, and mentoring activities [2], [3].
To support instructional designers and faculty in this endeavor, we have leveraged the APIs of OpenAI tools to create Transcriptto, a Python program that contains clever algorithms that aid in the crucial steps in lecture preparation, allowing instructional designers and faculty to have a better starting point when starting the development of an online course. Transcriptto utilizes a straightforward yet robust workflow, incorporating openly available technologies such as Pymovie, FFmpeg, OpenAI’s Whisper, and ChatGPT. It transforms video lectures into polished text, supporting various input types, including audio files, and pre-existing scripts. The tool enhances clarity, removes redundancy, student interventions, semester-specific details, and provides a user-friendly configuration file for easy customization without requiring programming knowledge. The resulting content is optimized for information processing and fine-tuned via the OpenAI API, yielding not just a polished transcription that may be used to record a streamlined lecture more efficiently, but a multifaceted educational suite that generates outlines for slides, quizzes, assignments, and discussion questions. All learning objects are based on the content being taught.
Preliminary test usage of Transcriptto indicates faculty are enthusiastic about options for enhancing existing lecture materials and creating an optimized experience for their online course. We estimate this process will contribute to cutting the time it takes to write a script for studio sessions, and faculty will also appreciate the efficient creation of high-quality additional learning objects to evaluate students and maintain engagement. To assess the efficacy of Transcriptto, we employ a modified schema after Merriam [4], conducting a survey and focus groups with faculty users who provided their content as input for the program to process. Our aim is to gather in-depth insights, experiences, strengths, weaknesses, and suggestions related to the tool, actively facilitating discussions, and analyzing qualitative data to draw conclusions about its efficacy. The tool’s adaptability, driven by ongoing feedback from faculty, instructional designers, and other experts in the field, ensures its alignment with real-world educational needs.
Keywords: AI, Instructional Design, Online Transition, Online Education, e-learning, OpenAI, Python, Time Saving Strategies for Faculty, Lecture Capture, Repurposing Video Lectures.

AU  - Gerry A Pedraza
AU  - Sunay Palsole
CY  - Portland, Oregon
DA  - 2024/06/23
PB  - ASEE Conferences
TI  - Board 49: Work in Progress: Using Generative AI for Reducing Faculty Workload in Online Engineering Courses
UR  - https://strategy.asee.org/47045
ER  -