AI in Aerospace Engineering

It seems imminent that AI is eventually going to be part of the workflow for most, if not all, aerospace engineering workflows. Some companies are reluctant to embrace the technological over fears of losing control of their proprietary data while individuals have concerns ranging from having the skills necessary to use AI to the AI outright eliminating their job. Another concern of AI is whether of not accuracy of the solutions that it produces is good enough for aerospace applications.

The concerns over AI are well founded, and must considered by individuals and companies alike. For individuals, viewing AI as an additional tool in their toolbox versus and adversary is advisable. AI requires a human to prompt it and its responses should be reviewed by engineers to ensure they are accurate and correct. On the part of an individual, this requires no less knowledge or skill than before, but may aid the engineer in arriving at a solution sooner. In this sense and analogy can be drawing between the integration of AI and the evolution of drafting boards to CAD machines. While the change from manual to electronic drawings has changed the ways engineers work, it has not eliminated the need for them, and in many ways has opened engineering up to new possibilities and work positions.

Which leads to AI from the perspective of a company. As AI starts to introduce new possibilities to engineers, companies that outright refuse the technology or fail to implement it properly, may be eroding themselves away to a company of yesteryear. There are not many (if any) aerospace companies today that function without computers and internet and AI currently has the momentum to be just as crucial to a business someday.

A good place to start with incorporating AI into individual and corporate use is to start using the technology. At the time of this writing, the following four AIs are among to most robust and advanced, and offer many basic features to individuals for free.

• chatGPT - https://chatgpt.com/

• Claude - https://claude.ai/

• Grok - https://x.ai/

• Gemini - https://gemini.google.com/

Since these four all run on their own servers, individuals and companies alike should be careful not to share or include any proprietary data. If companies or individuals want to use AI with proprietary data, setting up Ollama on a local server or individual company machines is recommended. Ollama and the models available to it are open source and not as robust at the four given commercial AIs. However, since it is running locally, any proprietary data being shared with it remains on the local server or machine, and remains within the company’s control (it doesn’t get sent to an external server).

Benny Aero uses commercial AIs when protection of data, confidential information or proprietary data is not in play. When proprietary data does enter into the picture, we have a local AI server that allows us to crunch through information like never before imagined. If you would like to learn how to set up and run local AI, check out our "How to Run and Install AI Locally" article.