With the H2C and its Vortek system, Bambu Lab has already greatly reduced one of the traditional bottlenecks of multi-material printing: tool and material switching at the printhead.
Vortek makes filament changes at the toolhead extremely fast and greatly reduces the need for long purge cycles. However, once tool changing becomes this fast, another bottleneck becomes much more noticeable:
the time required for filament to travel between the printer and the AMS units.
In practice, the printer can be ready for the next material long before the AMS has finished retracting the previous filament and feeding the next one through the PTFE path.
This means that part of the speed advantage provided by Vortek is still limited by the fact that filament loading and unloading remains essentially sequential.
I think a dedicated Smart AMS Hub with filament parking and predictive preloading could solve this.
It could bring the H2C significantly closer, in terms of material-change speed, to a true multi-toolhead or multi-extruder system, while retaining the flexibility and compactness of the AMS ecosystem.
The idea
I would like to suggest a new type of AMS accessory: a Smart AMS Hub designed specifically for filament parking, preloading and parallel AMS operation.
This would not simply be a larger Filament Track Switch.
The idea would be to connect several AMS units to a central intelligent hub, with one independent filament path / parking position for each AMS.
For example:
AMS 1 ───► [ Parking position 1 ] ─┐
AMS 2 ───► [ Parking position 2 ] ─┼──► Printer
AMS 3 ───► [ Parking position 3 ] ─┘
Each AMS could load one filament up to the hub and leave it parked there, close to the printer.
Why?
Currently, a significant amount of time during multi-material printing is spent unloading one filament and then waiting for the AMS to load the next filament through a long PTFE path.
This becomes particularly noticeable on the H2C because Vortek has made the toolhead-side part of the filament change so fast that the AMS travel time can become the dominant part of the whole material change.
With a Smart AMS Hub, these operations could be partially performed in parallel.
For example, suppose the print requires:
Red → Blue → Green → Yellow
and Red/Green are in AMS 1 while Blue/Yellow are in AMS 2.
While the printer is using Red from AMS 1, AMS 2 could already load Blue up to its parking position in the Smart Hub.
When the printer needs Blue:
-
Red only retracts as far as its parking position in the hub.
-
Blue, which is already waiting in another parking position, immediately feeds toward the toolhead.
-
While Blue is printing, AMS 1 can unload Red from its parking position and preload Green.
-
When Green is required, the same process happens in reverse, while AMS 2 prepares Yellow.
In other words, Bambu Studio/firmware could schedule filament usage across different AMS units whenever possible, so that one AMS is preparing the next material while another AMS is actively supplying the printer.
This would effectively create a filament-loading pipeline.
Intelligent scheduling
Bambu Studio already knows the complete sequence of filament changes before printing starts.
The software could therefore optimise which AMS supplies each material and which filament should be preloaded next.
When equivalent or duplicate filaments are available in several AMS units, it could even prefer the AMS that allows the next change to be prepared in parallel.
The objective would be:
keep one filament printing while another AMS prepares the next filament.
The slicer could even take the upcoming sequence of material changes into account when deciding which AMS should supply a filament.
For example, if two AMS units both contain the same material, the system could deliberately choose the AMS that leaves the other unit available to preload the next required filament.
This turns the AMS network from a passive storage system into an actively scheduled material delivery system.
Filament parking
An important part of the design would be that changing material should not require retracting the filament all the way back to its AMS.
The previous filament should only retract far enough to clear the common output path and remain parked inside or immediately before the Smart Hub.
This would dramatically reduce filament travel.
Instead of repeatedly moving filament through the entire PTFE path:
Toolhead ───────────────────────── AMS
most changes could happen over a much shorter distance:
Toolhead ───── Smart Hub
The longer AMS-to-Hub movement could then happen in the background while another filament is actively printing.
This is especially useful for Vortek
The H2C already has a system capable of very fast tool changes.
Because of this, reducing purge time and nozzle-change time alone eventually produces diminishing returns.
The next logical optimisation is therefore the material delivery path.
A Smart AMS Hub would complement Vortek by moving filament preparation away from the critical path of the tool change.
This is conceptually much closer to what happens in a multi-toolhead printer, where the next material is already physically available at the machine and does not need to travel a long distance before printing can resume.
A Smart AMS Hub would therefore allow the H2C to take much greater advantage of Vortek’s speed and could make real-world multi-material change times much closer to those of multi-toolhead systems, without requiring a separate complete printhead for every material.
Possible hardware
The product could have:
-
4, 6 or 8 AMS inputs.
-
Independent filament presence sensors for every input.
-
Defined filament parking positions.
-
Optional internal drive motors to assist filament movement.
-
Automatic identification of which filament is currently parked in each path.
-
Support for AMS 2 Pro, AMS HT and future AMS systems.
-
Communication with the printer so the firmware always knows which parking positions are occupied.
-
Sufficient independent paths so several filaments can remain staged simultaneously.
The important difference compared with a passive PTFE splitter or the current Filament Track Switch is that the hub would be designed around multiple simultaneously staged filaments.
I think something like an “AMS Smart Hub” or “AMS Nexus” designed around filament parking, parallel AMS operation and predictive preloading would be a very useful addition to the Bambu ecosystem.
