Automating plastic collection

Good morning!
First, awesome work and thank you, I need all this info here and the community. This community is nice because I won’t have to first sell the plastic problem and the fact that the problem is its own solution. I see a lot of groups doing lots, but lets face it, the output of garbage is of quantum magnitude in comparison.

We know exactly what do to, but we could never do it fast enough nor scale up our effort enough to win this in time. (Yes that is only my opinion…. lol )

This is where I came up wit project Uranus.
The project Uranus is to create a ( Get READY!! ) decentralized robotic automated networked infrastructure to collect, process and reuse the plastic.

Simply put, use garbage to create robots (as much garbage as possible, some parts would have to be bought.) uses these robots to go collect plastic, and feed your machines. Use machines to build more robots to collect and products to sell.

* The Plan
– Think small, scalable, decentralized, automated
– Find plastic
– Collect plastic
– Process into usable plastic
– Print products
– Sale products
– Scale up and out

* How to start
– Few people together, they are U.
– Each U plan and design first Replicator, Extruder, Collector with helps of other U’s.
– U will obviously design models to tackle your environment. (I don’t ocean of plastic beside me, I have fields. Your meat and vegetable are eating plastic too….)
– Share and Open,

* What after?
After having good model for automated replication and product creation, lets call it a Uranus station. We can setup model for automated online store.:
– Order online from list of printable object or upload your own.
– Automate the order, creation and shipping.

Where am I at with this?
Well i’ll start by building your machine so I can use my plastic to build my robots.
I found that there is sensor capable of detecting plastic, I will be starting to order parts to test next week. But I have an extremely restricted budget. Lucky for me most parts are cheap. And gets cheaper as I move along this project… lol


Super interesting! Can you link me the sensor which detects plastic?! I’d like to check it out and start experimenting here too. Where are you based?

@papa @andyn The vision system needs to have a hyperspectral camera combined with the sensors for detection as well as analysis. The camera needs to have a reference database that it can check against to determine which plastic it is so it can sort it. This is the system used by the big recycling companies for sorting on a conveyor. The kinect only gives a ir dot reference which will detect that it is there or it has moved but will not detect what plastic it is unless @papa you have solved this.
So to summarise the vision system comprises of a camera and sensors ( capacitive) to do the sorting function
Hope this helps

Any news? Update of your project?
Thanks DiB

To note also that I will start by building robot to gather parts that are easy to access and move toward more complicated scenario as progress allows it.

I like the idea of color contrast, I am playing with Kinect right now, I will be able to use those to test the detection process.

Yes, I think some sort of visual processing would be best. Simplest would be to identify objects that have colours that do not belong in that environment, like a red or blue bottlecap in a green field. More advanced, it could be trained to recognise objects it sees as being a beer can, a plastic water bottle, a leaf etc…

Yes you are correct, the robots will need to have cameras, and other sensors to, which will enable it to see the difference between a cucumber and plastic. The fact that something is not going to be easy does not mean it should not be done. And the fact that it has not been done does not mean it is not possible.
The easy solution was to throw it out.
I can point my hone at an object and it will give me all the info I need on it. IT is totally feasible to create machine that will find, recognize and collect plastic.

I don’t think you can actually use a capacitive sensor to differentiate between different types of plastic, if that’s what you were thinking. Or between plastic and natural materials like stone and dry wood. If it were that easy it would be done that way already. You might be able to use capacitive sensing in conjunction with other sensors to determine if something is NOT a piece of metal, green twig, vegetable, live animal etc. and thereby isolate potentially plastic items by a process of elimination.

I started a Facebook page for the Uranus project:

Thank you. They are called capacitive sensors:

I am located in Gravelbourg, Saskatchewan.
Right now I am generating interest in my local environment. This seems to be working I have schools that would like to join. The issue is always the same, $$$.
My plan is to gather interest, centralize communication, come up with robotic model and I was thinking of launching a crowdfunding project to get it started for the first year.
I have a friend which which does industrial machine design, waiting on a reply to see if he can help with the designs.
I will keep you updated, and please inform me if you have progress, idea or happen to know someone that is good at running a crowdfund…. (I’m good with robots).
At some point I want robots that goes on and in water, but I will have to start with my environment.
Thank you for your interest.