Using AI to improve the efficiency in the process of designing and launching a rocket.
Ever since I was a kid, I’ve loved the idea of space exploration. I’ve always wondered what is up there, above our blue sky, how do these billions of stars look like, and how does our universe work? But then once you start to grow up, like the teenage humans we are, we start to see flaws in some systems, and for the process of designing and launching a rocket, it’s the exact same thing! Let’s be honest, the process of launching rockets is really slow, it’s not the most efficient thing out there, but how do we really improve this system?…
Well, I’m not talking about improving Isp (specific impulse) or talking about revolutionizing rocket propulsion (ie. Antimatter, Ion propulsion, etc), what I’m talking about is integrating generative design to improve efficiency in designing and launching a rocket.
Motivation 🚀
Like I mentioned above, I wanted to figure out an efficient method to launch and design rockets because I want to see increased frequencies of rocket launches and push the boundaries for space exploration! If we can achieve this, we will reduce the cost/per launch to make space exploration much more accessible to companies and people to also make the possibility of small space startups to succeed.
Generative Design 🏡
First, before I actually talk about more of my proposal idea, let’s talk about generative design and how it works and the intersection of generative design and additive manufacturing I used in my project!
What is generative design?
Well, in simple terms, generative design is a technique used in manufacturing products. Generative design leverages the power of AI to create 100s or if not 1000s of possible solutions (different CAD models)for something to manufacture! First of all, we design the initial CAD model for our generative design model, then we input the parameters we want for the generative model to create the generated extra solutions. Next, the generative design algorithm, leveraging AI, would generate new solutions to the CAD model we created. We can have as many solutions as we want, although of course computer hardware and processing speed. Next, you and your team would assess the viability of each of the models to find the perfect model that matches the requirements of your design problem 🚀!
Below is a diagram I created on an app known as Lucidchart, it’s a great app to create process flowcharts which is what I did. Alongside that, you can also create team workflows and database diagrams!
Additive Manufacturing (3D printing) 🧱
Once we finalize our CAD model that our generative design model created, we can then use different manufacturing techniques such as 3D printing we can use the final model we selected, then we can make our online model into an actual physical object that meets the requirements and solves our design dilemma. We can also integrate AI in the 3D-Printing process to also streamline the process of creating our product after the generative design model.
Let me explain to you some of the benefits of integrating AI in the 3D-printing process:
- First of all, we can fully automate the 3D-printing process, this means that the AI can design the model via leveraging generative design from importing it into the slicing software to then 3D-print it to create our model, it can overall increase efficiency and produce our product and market it to consumers much quicker.
- Next, we can expand the range of materials that we can use in 3D-printing, this would be ideal in an industry such as the aerospace industry. A prime example of this is the futureAM project launched by Fraunhofer Institute IWS in 2017. The company had explained : “Aircraft engines could operate at higher temperatures and more efficiently if most materials did not fail at temperatures above 1200 degrees.” Although how is this company working on integrating AI to generate a wider range of materials? It requires a lot of monitoring and special requirements, we can use AI to monitor the material being made via the 3D-printing process to ensure it will be a stable material while in use.
- Finally, we can use AI to simply optimize the overall process, AI can be used to check the manufacturability of the parts in products alongside preventing printing errors in the process.
Generative Design + 3D-Printing + The Aerospace Industry
Well, when we look at implementing generative design in the aerospace industry, the overall process doesn’t really change at all! We first design the model in a software such as Autodesk’s Fusion 360. Next, we would configure the parameters and generate the different CAD rocket models that would help us overcome our design problem. Below is another process flowchart I made that explains how the generative design process works when looking at rocketry.
Next, we can 3D-print the rocket to reduce the time to make the rocket from 24–48 months to just a few weeks. A company working on 3D-printing rockets, known as Relativity Space, claims that it can reduce the cost to make a full-sized rocket from 100 million dollars to just 10 million! Using the intersection of these 2 technologies we can reduce the cost of space travel and make the frequencies of rocket launches much more efficient alongside hopefully make space travel much more accessible to smaller aerospace startups.
Socials + Conclusion 👋
Well, that is it for this article, I really hoped you enjoyed it and learned about something new since many people don’t know about generative design and also the intersection between generative design and 3D-printing. If you enjoyed this article and found it insightful, I would really appreciate if you shared it with someone who you think would benefit from reading it and giving it a clap!
Socials :
