Inspiration

Expand 3d printer application to scrap and repairs.Its high time to use things resourcefully by waste management for better ecosystem.Most of the things are converted to scrap due to minor breakages and lack of resources to fix them though there are possibilities to nurse back to their shape.Just like worms REGENERATE itself by regrowing broken parts wrecked things are reconstructed.Aim is to bridge the 3d object processing and 3d printing technology for Revival of things to increase the life of objects and for reducing wastes.

What it does

Generates the broken or missing part of object from 3d image of object.This object can be used to reconstruct the thing by generation of broken part from 3d printer.If for example a chair leg was broken or cracked to 3/4 of its length instead of throwing it away it can be fixed by 3d printing missing part of chair and joining by adhesives.

How I built it

3d ModelNet dataset trained using tensorflow2 implementing 3d DCGAN for generating a broke object's missing component as other object.This object can be re scaled with the support of 3d Kinect to generate compatible object to the main object for 3d printing thereby making it reusable and reducing scrap.

Challenges I ran into

Migrating to tensorflow2,tensorlayer2. Implementing DCGAN on3d objects.

Accomplishments that I'm proud of

What I learned

What's next for FixIt

Integrating with 3D kinect to develop end to end application for 3d printer which can be used from household to industries to waste management.

Built With

  • deeplearning
  • modelnet40
  • tensorflow2
  • vae-gan
Share this project:

Updates