Yes, working on it!
From a user perspective, let me raise a possible issue: the filament is exposed. Is it possible to have a simple sealed space for moisture protection? (Adding acrylic into the slots of the profile and inserting silicone might be a good direction, but this is not a core function.)
Youâre a genius. Has Snapmaker contacted you yet? 
Please keep us updated. Iâm already curious how far the project will go.
A chamber for the filaments that also draws out moisture would be the icing on the cake.
I have transparent CMYKOGVW (extended color gamut) , just waiting to test it out!
I have a sidequest thatâs almost done that will allow much more freedom in that regard!
Great job. Well done on the sheer dedication to purpose!
How do we follow up on this?
Are you launching on kickstarter or collaborating with snapmaker maybe?
You have ticked off almost all the features I had in mind for an ideal tool changer + AMS setup. The only major thing missing that I would like to implement is a way to prime while the tool is docked. Something along the lines of replacing the rubber ooze stopper with a small motorised rotating cylinder with an appropriate easy release surface and a scraper on the bottom side. This should allow purging and priming in the same sequence, as long as it was pulsed rather than one continuous extrusion - think grains of rice âpoopâ rather than noodles.
You mentioned somewhere that you initially tried a setup where all filaments could be loaded into any toolhead - do you have any photos? Do you think users could mod your kits to allow this without too much trouble? I think this would be very useful in a mixed nozzle size setup, with fine nozzles for outer walls or full spectrum colour blending and details, and a large nozzle for inner walls and infill or strong mechanical prints and worth any PTFE spaghetti.
I know I am long in the tooth when writing posts so you probably didnât want to read all of the first two posts I made (I donât blame you 
) but if you read them (and click the youtube video links) you will see most of your questions have been answered and it already does most of what you suggest. I do not have photos of the filament rats next from the active hub endeavor, that would take a lot of fiddling to make reliable and also slow the system down as a whole since the filaments would have longer travel paths, but itâs not impossible to do hardware or firmware wise.
I saw and read that you can purge while the tool is docked, which is allready a great time saver, but to clarify my last post - I always envisioned removing the need for the prime tower also with something along the lines of the setup I described above, not because of the waste of filament, but because of the waste of time. This way the total time for a tool change could be reduced to a second or so, which obviously can make a huge difference in print time for a part with many tool changes, making it more competitive with single filament prints.
A rotating wheel with an appropriate print surface, kind of like a really small belt printer but with a cylindrical surface of a wheel 0.2mm or so under the nozzle. Extruding while rotating the wheel should do a similar job as a purge line, negating the need for a prime tower. If the wheel were slightly cam-shaped, one area could still be used to block off the nozzle if desired. The design and tweaking would be slightly more complicated than what you already have, though it would use the same amount of IO using a 360 degree servo. if the whole purge and prime cycle is done in short 5mm long purge lines, then the âpoopâ would proabably take less space too.
I think thatâs overly complicated for what you are trying to achieve, if you want a short prime before the toolhead is picked up you can do that using the current wiper system the sidecar has, a purge is basically a really long prime. Small bits of filament are pretty hard to direct where to go as well, they kind of fly wherever they want, especially when static charge is involved, even the short bit of filament thatâs produced during the unload tip tuning process has a mind of itâs own and I had to change the wipe pattern to tame what it did between unloading and loading. Iâve been running no prime tower with just an extra wipe on the silicone brush on tool pick and itâs been working pretty good, that is another option.
Are you able to time purges just-in-time for a tool change, so the filament in the hot end hasnât degraded or oozed, then a quick nozzle wipe before printing? Iâd imagine that gets most of the benefit of a prime tower with very little waste.
You can run any extruder whenever you want, so technically yes, but the more actions are you are doing at once the more potential race/gating issues you might encounter. The U1 is sensitive to stepcompress errors. Lots of people run no prime tower at all with flush to infill and/or inside/outside slicer settings, and depending on the dwell time between toolhead use, it might not be needed at all anyways, degradation wise.
All very well, but how could that be implemented as an add-on to the U1? Itâs just not practicable.
Mechanically I donât believe it would be any more complicated to implement than the system Sidecar already implements.
Sidecar makes a good point regarding the necessity versus just not using a prime tower and priming to infill though priming to infill is not always an option, and ideally a tool should be printing as close to perfect as possible from its first extrusion after a tool change
To further minimise tool change time, it would be great if the nozzle wiping also happened while the tool is docked or as part of the undocking move.
Please explain. So far as I can see, it would involve re-engineering the U1 rather than simply adding an external unit. If starting from scratch, you might be right â but even if you could fit the necessary mechanics into the available space (and find a way to connect them up), retro-fitting an existing U1 would be for the brave.
It would only involve re-engineering the tool wiper part with a motor or servo, much like Sidecar has already done, but in a different orientation. You can see the added servo and docked purge routine in the side car videos.
So far as I know, Sidecar has not re-engineered anything mechanical within the U1 cabinet, just added code for driving the external hardware. What do you say, @Sidecar ??
It can wipe right before pickup, it also gets wiped by the silicone brush as it comes out of the dock, so itâs pretty clean when it has to start printing. Mechanically a drum/scraper system wouldnât be impossible, but again, I think the tiny bits of filament would be the problem. Like when you have a failed print and have to scrape tiny letters/bits off of the built plate, they go everywhere. I think it would take a lot of testing and material choices to get a drum/scraper system working, but not impossible.