ARTIFICIAL INTELLIGENCE ChatGPT for power electronics industry: Hit or miss?

Related Vendors

CeramTec GmbH

GVA Leistungselektronik GmbH

Koki Deutschland Niederlassung KOKI Europe A/S

OpenAI’s ChatGPT has taken the world by storm to generate human-like responses. The AI tool assists professionals in content creation, research, and even coding in various programming languages. But little did you know ChatGPT tries to mimic an electronic engineer’s response. Let us look at the impact of ChatGPT on the Power Electronics Industry and evaluate its assistance to the engineers.

The power electronics industry incorporates coding, design, modeling, simulation, tests, and analysis much more than content creation. Contrary to popular belief, ChatGPT isn’t limited to content creation as the software writes code, troubleshoots it, performs complex calculations, and unloads various low-value tasks in the power industry.

1. Guidance/training to beginners

ChatGPT is a great tool to train beginners, hobbyists, and students in the power electronics industry. The software can generate mass content within minutes. However, the content needs extensive attention to editing and fact-checking for obtaining the best results.

2. Generates code

One of the best features of ChatGPT is to write code in various programming languages. Since hardware-level programming is a bit of a challenging process, the software enables you to write complex code multiple times.

ChatGPT writes code in extensive programming languages used in power electronics such as C, Embedded C, C++, Python, and HDLs like Verilog, VHDL, etc. Moreover, ChatGPT can take “code input” to find an error and fix it.

3. Solves complex equations

ChatGPT tries to provide steps and necessary guidance required for solving complex mathematical equations in power electronics. It provides solutions for Boolean expressions, calculus, and popular signal-processing algorithms such as FFT (Fast Fourier Transform).

4. Enhances designs

ChatGPT provides extensive information for designing various electronic devices and PCBs. It generates adequate information regarding design rule checking. The powerful AI tool takes direct inputs for real-life design problems with many values and instructions. Moreover, ChatGPT provides solutions for regulatory compliance in the power electronics industry.

5. Technical support to customers/clients

Many companies including the power electronics industry, are using ChatGPT as a bot to generate responses to customers/clients. Due to its exceptional human-like responses, ChatGPT offers better solutions to the user compared to VoIP-based pre-programmed caller responses. The software enables the industry to reduce these low-level tasks of providing accurate responses and saving money.

ELECTRONIC DESIGN AUTOMATION

Five EDA tools to empower your next PCB design

Can ChatGPT beat an electronics engineer?

No, ChatGPT cannot beat an electronics engineer. Figure 1 explains why.

1. Inability to integrate with other tools

One of the greatest limitations of ChatGPT is that it cannot integrate with other software tools for designing and coding purposes. For better understanding, ChatGPT can’t perform operations on any other software. It can only suggest responses, which a human has to implement on the software/platform.

2. Lacks innovation

ChatGPT responses are very generic and less creative in nature. The responses tend to be less conceptual and are heavily dependent upon training results that have been gathered through the Internet. The AI tool fails to correlate, research, and analyze advanced questions like a human mind.

3. Limited capabilities for the design and analyze

ChatGPT cannot perform circuit simulations like an electronics engineer. The AI software can only provide information related to the software, components, and manufacturers.

When design prompts are made in ChatGPT, the software tends to provide generic solutions to the problems. At times, the text seems to be correct because it is coherent. But engineers have been pointing out various errors and failures to answer the optimal solution.

4. Failure to recognize the past set of values

Certain problems are quite long and require multiple steps that may be interdependent upon the past set of conditions. ChatGPT often fails to provide continuous responses depending on previous input data. Many times the software provides a completely different response for a question dependent upon previous inputs.

5. Untrustworthy evaluation

ChatGPT troubleshoots code, finds errors in text, and fact-checks information. But it cannot always be correct. ChatGPT fails to take long lines of code and generate responses for it. Many times, the software has been noticed to provide incorrect information and generate non-existent reference links in its response to support the data.

Conclusion

Despite various limiting factors, ChatGPT can assist an electronics engineer. The software is capable of performing low-level research, fact-checking, coding, designing, and creating content. In conclusion, ChatGPT is a powerful AI tool that can be an electronics engineer’s friend.

(ID:49536380)