3D Printing G-Code Gets an Upgrade: T-Code

Good old G-Code still manages many 3D printers, great and small. Just like the STL, it’s a standard that enables collaboration while also holding the additive manufacturing (AM) industry back. Johns Hopkins University researchers Sarah Propst and Jochen Mueller believe we may need an upgrade. In a paper published in Nature, they argue for the creation of T-Code.

Through their work with Direct Ink Writing, the duo has recognized the limitations of G-Code. Modern nozzles can switch between materials, mix them in specific ways and proportions, adjust extrusion profiles, and even cure selected areas of parts. In G-Code, print movement is paused for the print head and other actions to take place. When this happens, the print may fail. Issues such as over- or under-extrusion, pressure buildup, premature hardening, adhesion problems, and layer profile changes can occur.

To address these challenges and allow for deeper integration of future functionalities, they propose Time Code. Instead of viewing machine instructions as a single uninterrupted path, they envision a system akin to how software like Ableton Live organizes music. In such tools, a track runs while the arrangement view displays different music sources, instruments, loops, beats, and vocals, each triggered at specific times or in parallel. Modifiers or additions can be individually adjusted, timed, and rearranged in real time.

Rather than writing one static code, Time Code would allow for the programming of modifiers, unique time-based structures, and real-time adjustments. Instead of merely moving the toolhead, it could orchestrate order and disorder across four dimensions. If we take inspiration from Ableton and similar software, we could shift our focus toward developing tools that interact and enhance each other, rather than merely printing in a linear, rigid manner.

Imagine being able to design software “instruments” optimized for specific nozzle extrusions. Sally could then create various effects—code snippets designed for particular outcomes—while others work on triggering events or dynamically altering parameters. If 3D printing operated in this way, wouldn’t the technology become exponentially more powerful?

The team has explored implementing their technology specifically for multilateral printing on several printers. They believe that linear advance or other pressure prediction and management solutions will not suffice, particularly for Direct Ink Writing. They also argue that at much higher resolutions and with more materials, the sheer number of commands would cause excessive lag. Additionally, they maintain that linear advance must be fine-tuned for each specific printer, material, or process, whereas a broader approach could be more efficient.

In their model, G-Code still exists but is used exclusively for head movement. A list is generated based on a timestamp-filled “velocity profile,” which is synchronized with the G-Code. At specific timestamps, events are triggered accordingly. The team uses Python to integrate these elements, pinging the printer a few times to synchronize the script with the machine and performing additional pings to maintain synchronization.

The team goes on to create gradient filaments and gradient parts using coextrusion at a constant print speed. In multi-toolhead systems and cell-based manufacturing, their approach could significantly reduce time and complexity. Looking ahead to the future of additive manufacturing, we are moving toward voxel-level resolution. At the same time, individual voxels, clusters of voxels, or voxels composed of different materials will exhibit distinct properties. Nozzles will change, materials will switch more rapidly, and a greater variety of materials will be used.

This increasing complexity, particularly in bioprinting and printed electronics, could become overwhelming. Until now, file-based approaches like those used in 3MF were considered sufficient for describing complex structures with greater accuracy. However, T-Code could offer a novel, creative way to conceive, develop, and scale the manufacturing of functional geometries that far surpass what is currently achievable.

I love T-Code as a concept, and I think we need to explore ideas like the remix approach or T-Code itself. As we rethink how things are made, we should remain mindful of how we approach this transformation. However, beyond the concept, will this truly help people, and is it the best way to do so?

One has to wonder whether much of this could be addressed on the firmware side or by upgrading hardware, such as the control board. A friend of mine immediately suggested that a PLC might be a simpler way to solve this problem. The drift between the two synchronized systems created by T-Code seems like a complex challenge to manage. So, while I’m captivated by the idea, I’m also concerned that it may not ultimately provide a practical solution.