That sounds like it could be very useful.

There are a few examples of converting plastic bottles to plastic string

e.g. https://www.youtube.com/watch?v=tywJmJK7rJA

I have a friend that teaches technology, including 3D printing.  Part of their class projects is printing and finishing low-cost plastic prosthetics made to fit a specific child in their community.  Their greatest expense and frustration is getting the filament.  By being able to use a DIY filament production, they could do so much more, as well as help more people.

Their primary effort is producing artificial hands for children with Amniotic Band Syndrome/ADAM Complex, which often results in partial amputation of the hand, or foot.  Statistically speaking, 1 in every 1200 births involves Amniotic Band Syndrome.  In 2014, that average was 3 children, every 5 minutes (36 children every hour), somewhere in the world.  So far, they have not progressed into designing them, but are using open-source plans.

Bringing this type of support sytem to outlying communities would be greatly simplified, and costs greatly reduced.  In the process, cleaning up the environment of plastic waste would greatly benefit the ecosystems around the world, with recycling on a local level.  Imagine a 'trash-drive' to collect plastics, every few years, for a new prosthetic limb to fit each of those children affected, while the previous one gets refurbished and refitted for another needy child.

While I am not directly involved, I feel it important enough to try to find a way to improve the world around me, if only by having an idea that can inspire others to action, where I cannot afford to provide such life-changing technology, myself.

I think one of the issues here might be the different types of plastics.

I've not got into 3D printing yet myself but it appears that the filaments used tend to be either:

PLA (Polyactic acid)

ABS (Acrylonitrile butadiene styrene)

PVA (Polyvinyl alcohol)

Whereas the identifiable recyclable household materials (soda bottles, milk jugs, cd cases, plastic bags) tend to be:

PP (Polypropylene)

PE (Polyethelyne)

PET (Polyethelyne terephthalate)

PS (Polystyrene)

V (Polyvinal chloride)

Milk jugs and 'disposable' plastic shopping bags, generally are made HDPE.  Same plastic, different process of manufacturing, for different end products.  The hand in the second article is HDPE.

Alternatively, PLA is a plastic derived from corn, frequently found as disposable plastic shopping bags, and is commonly used in making the 'robohand' prosthetics.

Got Milk? Then You’ve Got 3D Printer Filament at a 99.7% Discount | 3DPrint.com

Looks as if for the HDPE milk bottles one could even use something like a deep fat fryer as a melting pot

Making Blocks out of HDPE milk jugs - All

and then drain that into the extrusion unit

http://reprap.org/wiki/Recyclebot

Temperature control, extrusion speed control and take up spool speed control could be controlled with a microcontroller.

The difficult thing to control is probably contamination of the recycled materials - stuff like labels, glue, foil, milk residue etc.

I've been working nearly four years (off and on) on an improved design of Ben's Briefcase 3D Printer. One of my intentions was to include (or design separately) a folding filament making station. (about the same size) I've since abandoned the idea after finding out a few problems.

1. Assembly - From the motor to the hopper to the feed screw to the heating chamber and finally the extruder head, It all needs to be tight fitting and of solid construction. Once everything is assembled, it's would need to stay assembled. Otherwise you'd need to carry along tools (large tools) to put the parts back together. You don't want to have to assemble and disassemble it everywhere you went. It would take too long.

2. Size - You can leave everything assembled and just make it so that your winding spool folds down and take off your hopper, but your overall length is still determined by the length of the unit itself. This would just be making a box around it with a handle on top and would be very awkward to carry. (You may just put a carry handle on the top of the Filabot Wee.)

3. Convenience - Would it really be all that useful to make it portable? To even start making the filament you'd want to have a collection of bottles saved up. It seems easier to transport a few bottles to a local MakerSpace rather than lugging around a large extruder. Even still, the bottle need to be prepared ahead of time. This means removing the stickers and separating the caps and the tamper proof ring from the bottle (Usually a different type of plastic). The bottles would probably have to be boiled first as well.

It's nice to be able to pack it away when you're not using it, all nice and neat like, though. But that's no different than my sewing machine having a cover with a handle on it.

Even if you wanted to make it so that you disassembled everything. Ben would only be designing a toolbox to carry the parts and tools in. I'm sure it would be an awesome box, but where is the "hack"?

I'd love to see this project too. Maybe Ben can find a solution to the troubles I've found that has kept me from making it.

I agree, it could be difficult, and there are others making extruders.  None that I know of grind, melt, extrude and have a take-up reel, in one compact unit.

Not many maker-Spaces around, much less those with extruders.  How many in your town?  None in mine, and only 3 listed for the entire state of Texas...  Nearest to me is UT Austin and they have 3 different 3D printers listed, no extruders, whatsoever.

Sure, someone has to collect, and clean the jugs and/or bags, let them dry and run them through a heavy-duty shredder.  It's called supplies and preparation.  Can't think of a single construction project that doesn't need supplies and preparation.

Why does it need a bunch of heavy tools?  Why not have interchangeable heads, for different plastics?  For every problem, there is a solution.  It may not be an obvious solution, but where there's a will, there is a way.

Where's the 'hack'...?  How about a completely portable 3D Printing Factory, capable of being carry-on luggage, for air travel?  Isn't that worthy enough?  I know there's a need for it, and I know people that travel internationally, on educational projects, even school trips...  Why not a complete system capable of going to Central America, the Middle East, Africa, India, Asia, or even the next city over, to make 3D prosthetics, for people that need them?

How about starting off with a two-step process ?

First step to create a uniform stock material from the recyclables. Second step to extrude from that newly formed stock material into a spool of filament.

Something similar to the concept of glue-gun sticks, with the intermediate stock being made using a basic melt and mould type process.

IMHO, that requires an unnecessary melt, more energy, time, and an additional molding process.  Might as well make pellets, at that point.  That's really the purpose of using shredded stock, then being able to combine the final grinder, melt and extrusion in one portable unit.  The grinder wouldn't be necessary, if all shredders produced the same size particles.  However, there are other ways to do it.  Using pre-shredded stock and a decent sized fold-out hopper, then keeping the thing fed, producing the filament should be relatively easy.  At the same time, it may be possible to use available pellets (I've found them in bags and drums at salvage yards, from time to time), as well as tail-ends from other reels, since they would also go through the grinder.

Maybe temps could be controlled with a temperature sensor and PWM power supply?

Suppose the whole case-top opens, like a doctors' bag, to be the feed hopper, then that ends in a grinder/auger, at one bottom corner, which feeds the heater-head/extruder assy., and out, past cooling fans, to the take-up reel.  Might be able to include room for a selection of heads, for common types of plastics, so as not to need cleaning.  People leave old glue-sticks in the gun, and just feed a new one in, as needed, why not with the extruder, and even the printer, as well? 

Take-up reel could fit on a motorized, fold-out arm, like a movie projector.  Store one or more reels, depending on size, inside the hopper, when closing down for transport, or storage.  Make all reels so they are interchangable, from take-up, to being used in the printer.

Should be a relatively easy experimental process to figure the extrusion speed for consistent filament, and use a stepper motor at a relatively controlled tension.  Once temperatures and speeds are determined, lock in the settings, so they can be selected by a simple controller, similar to the one for the existing printer.

Out of all of it, the hopper and take-up reels would require the most space.  Combining reel storage and an empty hopper is almost a no-brainer.  Maybe arrive with several spools full of the expected plastic filament, and be using them, while extruding more?  At least it shows a continuity, but still encourages recycling the plastic, rather than having the person doing the printing supply everything.  Maybe they show up with the twin carry-ons, and set up the extruder, then, once it's filling an empty reel, take a pre-filled reel, and start the print.  As parts come off the printer for cleanup and assembly, the whole process continues.  Once the last piece is printed, the extrusion continues until is has caught up with restocking.

Never having done, or personally observed either, I don't really know which process is faster, but, from what I gather, the printing is faster using filament, than extruding produces it.  The extruder would need to be more robust, in order to provide filament for multiple 3D printers, for classes, but by having a uniform and ready means to supply filament, the students get more bang for the buck, out of the classes, as well.