Automated Instantiation of Designed Experiments with Additive Manufacturing

C#
Python
3D Printing
Software Engineering
Automation

Project members and contact information Paul Lee: leep3@oregonstate.edu Devin Swift: ds@devinswift.dev Erin Villasenor: villaser@oregonstate.edu We worked with our project partner, Will Allen of HP Labs, to simplify and automate the process of creating multiple objects in SolidWorks at once, each with different measurements. Our goal was to reduce the tedium of having to hand-design each individual object. We successfully simplified the process of creating new randomized SolidWorks objects. We also created a learning algorithm which can take performances and create more objects that perform better relative to the data given. We tested our process by using a launcher and propeller model, where we randomized the creation of 16 different propellers, and tested each of them using our launcher. We then recorded the results of which ones had the most time spent in the air from launch to when it first hits the ground. After we recorded our data, we were able to insert the results back into our program and refine our propellers so that when randomized, they would be more similar to the ones that performed better, slowly improving over the course of different generations until we could find a model that would be closest to perfect. Below are screenshots that include: 1. A sample of our Graphical User Interface (GUI) 2. Our launcher and one of many propellers used 3. How our program looks in SolidWorks. 4. A .csv file of the data created after running our program 5. The data we recorded from our first generation of propellers (Where the entry in blue is the worst performing model and the entry in red is the best, as well as the total average found in the bottom right) 6. The data we recorded from our second generation of propellers (Where the entry in blue is the worst performing model and the entry in red is the best, as well as the total average found in the bottom right)

0 Lifts 

Artifacts

Name Description
Team GitHub link A link to our project's GitHub, where all code is stored   Link
Project Demo Video A YouTube link that provides an overview of our project, as well as a demonstration of it in action   Link
Requirement Document The requirements that we built our project around from the beginning of the year   Link