Related Experiment Video
Updated: Jul 23, 2025
16:05
Using Micro-Electro-Mechanical Systems MEMS to Develop Diagnostic Tools
Published on: October 1, 2007
7.6K
Can ChatGPT Help in Electronics Research and Development? A Case Study with Applied Sensors.
Zoltán Tafferner1, Illés Balázs1,2, Olivér Krammer1
1Department of Electronics Technology, Faculty of Electrical Engineering and Informatics, Budapest University of Technology and Economics, H-1111 Budapest, Hungary.
Sensors (Basel, Switzerland)
|July 11, 2023
Summary
This study evaluated ChatGPT for electronics R&D, finding it offers good controller suggestions but flawed sensor recommendations and fabricated citations. Its use in embedded systems development requires careful verification by researchers.
Area of Science:
- Electronics Research and Development
- Artificial Intelligence Applications
- Embedded Electronic Systems
Background:
- The integration of Artificial Intelligence (AI) tools like ChatGPT into scientific research and development is rapidly evolving.
- The application of AI in specialized fields such as electronics R&D, particularly for embedded systems, remains under-explored.
- A gap exists in understanding the practical capabilities and limitations of AI in generating technical specifications and conducting literature surveys for electronics projects.
Purpose of the Study:
- To investigate the applicability and performance of ChatGPT in electronics research and development.
- To assess ChatGPT's ability to provide accurate information on controllers, sensors, and design flows for embedded electronic systems.
- To evaluate the reliability of ChatGPT in performing literature surveys and identifying relevant scientific papers.
Main Methods:
- A case study approach was employed, focusing on applied sensors in embedded electronic systems for a smart home project.
- Specific prompts were given to ChatGPT regarding central processing controller units, sensor selection, hardware/software design, and literature surveys.
- Qualitative and performance analyses were conducted on the generated responses, including specifications, code, and citation accuracy.
Main Results:
- ChatGPT provided acceptable recommendations for central processing controller units.
- Suggested sensor units, hardware/software design details, and generated code were only partially accurate, containing occasional errors.
- The literature survey function produced non-acceptable, fabricated citations, including fake author lists, titles, journal details, and DOIs.
Conclusions:
- ChatGPT demonstrates potential in assisting with initial electronics development tasks, particularly controller selection.
- Significant limitations exist regarding the accuracy of sensor specifications, hardware/software design recommendations, and code generation.
- The AI's tendency to fabricate citations necessitates rigorous verification of all generated scientific references.
