Inspiration: We are a team of mechanical engineering majors and physics minors. Informed by both our studies as well as tangible industry experience watching the product design process take form at companies like Apple, we know firsthand that one of the biggest challenges facing the product design process is the disconnect that exists between industrial designers and mechanical engineers. While possessing similar skillsets in some respects, their means of implementation can, at times, be strikingly different. We are inspired to bridge the gap.
What it does: We have built a ML URL extension that takes in 2D industrial design sketches and outputs a color-coded image, identifying areas that may be structurally vulnerable or parts containing complex geometries.
How we built it: This extension was implemented in python and trained via PyTorch by using industrial design sketches of consumer electronics.
Challenges we ran into: To provide an explanation of the potential issues in the design, we tried to implement Gemini API and integrate it into python, but it did not work as expected. As mechanical engineers, we realistically need more in-depth computer science knowledge in this area. We also need to further train our model to improve its accuracy.
Accomplishments that we're proud of: Proposing a realistic solution to an extremely crucial problem faced in physical product design.
What we learned: We learned the importance of fleshing out a solid roadmap for implementation prior to coding, as execution is always more difficult than expected. Additionally, harnessing the strengths of each team member is also very important.
What's next for Elevate: Further train the model to improve its accuracy.
Log in or sign up for Devpost to join the conversation.