The goal of this prototype was to create a recycling can that can automate the separation of trash and recycling so the user doesn't have to.
This model was the first prototype that was manufactured for the Eleven58 startup company. The models frame was assembled out of wood, aluminum, and some 3-D printed parts. The S57 operates using a conveyor belt mechanism and uses machine learning to identify and separate recognized trash items from recognized recyclable items.
For the S57, I conceptualized and designed the preliminary models for the systems early stages, as well as lead the manufacturing and assembly of our final S57 prototype.
The model shows the basic movement of an item placed towards the right end of the screen. In this CAD, the belt is yellow and the walls that separate the trash are gray and located on top of the belt.
First the item gets placed onto the belt.
A camera detects the type of object placed and sends a signal to start the belt.
The cameras detection controls whether the first wall moves or not.
We decided that this model should only rotate the wall to unobstruct the pathway if the material is recyclable. This way, in any case, only recyclable items are placed in the second bin.
To manufacture the S57, I worked with another engineer to machine and assemble as much spare stock as possible into a working prototype. The first model was made of spare wood and spare aluminum to create the frame, which holds the belt and the sorting mechanism. The general process goes as shown:
First we took 2 pieces of 48" * 1" * .5" wood stock and 2 pieces of 24" * 1" *.5" woods stock to create the general dimensions of the frames exterior.
We 3-D printed parts to hold the bearings which hold the shaft attaching to the also 3-D printed roller, which holds the rubber belt.
We then cut the rubber to 91" * 8".
Then we attached four small pieces of wood to hold the belt, motors and other 3-D printed parts.
After testing the belt, we placed a 4ft rod in the middle of the frame and attached the sorting mechanism, which included a stepper motor and a swiper.
We finally programmed the stepper motors to move the belt halfway, swipe a material, and move the swiper back to the original position.
After getting the frame to work, we cut the exterior of the body using wood.
The design I had created had not been manufactured, but instead was used as a guide to help assemble the current prototype. The errand we were given had to change when faced with the manufacturing portion because the funds to create a more expensive prototype were not provided. So we worked with the material we had available. This reduced costs by around 75% since the prototype with a metal frame would have costed upwards around $200- $250 while the actual prototype using spare parts cost 50$. The picture to the right displays only the belt and the frame.
Belt, Frame and 3-D printed parts.
Stepper motor and 3-D printed mount.
Roller (Blue), and bearing holders (Red)
This is our first trial. We were testing if the belt would move with one stepper motor.
This is a demonstration of what would happen when a recyclable is placed in S57
This is a demonstration of what would happen when trash is placed in S57
This is a video of me using a circular power saw to cut down wood for the S57 exterior
The model above is the most refined CAD that had been created in order to satisfy our goal with the constraints placed, but this was not the only model I had created.
over the course of my time in Eleven58, we had brainstormed multiple different designs that would do the same function of sorting trash into recycling. The model to the right was created after the model shown above in order to sort trash into three separate bins, but was never used because the we decided to work on a simpler model for our first project.
This CAD model is the very first model I fabricated during my time in Eleven58. The idea is the same, but the design is not refined nor made for manufacturing processes. It was just a model for general ideas
As of now, for the S57, I successfully designed all of the models and manufactured and assembled the frame and exterior with the help of another engineer.
The Good: The bin was successfully created and was able to move a belt, and the sorting mechanism when the camera was given 1 identified piece of trash and 1 identified piece of recyclable material.
The Bad: The bins recognition database only allows to The problems we ran into regarding the manufacturing were due to lack of quality stock which forced us to use warped wood in turn throwing off the precision required for the roller to roll continuously without having the belt bunching.