Digital Cuisine

Background

Before I began leading this project as a first-year PhD student at Columbia’s Creative Machines Lab, the development of 3D food printing (3DFP) had already been underway for nearly a decade. My advisor, Hod Lipson, had pioneered one of the first open-source 3D printers, Fab@Home, during his time at Cornell.

Initially, early experiments used peanut butter and cookie dough as test materials, but researchers soon realized that 3D printing could be used to create healthier foods, allowing for precise control over nutritional content. Imagine being able to adjust the amount of sugar, calcium, or protein in a meal based on your personal health data—this opened up a world of possibilities!

Technology

Additive manufacturing—the process of crafting 3D objects layer by layer—is already widely used in plastics and metals. But chefs have been "additively manufacturing" food for centuries—layering ingredients, flavors, and textures to create culinary masterpieces.

3D food printing (3DFP) applies this same layering principle but automates it with precision and customization. By leveraging data-driven health approaches, 3DFP allows for personalized, nutrient-rich meals at the press of a button. Combine the assembly capabilities of 3DFP with the high-fidelity heating and spatial control of lasers, and you’ll have a digital chef capable of localizing flavors, textures, and experiences with millimeter resolution.

Market Research

Because we envisioned this as an at-home kitchen appliance, we needed to understand our ideal customers.

In summer 2020 (peak pandemic), I participated in Innovation Corps (I-Corps), a National Science Foundation (NSF)-led training program. Through friends, family, and referrals, I interviewed 32 people, focusing on those with dietary restrictions and selective eating habits.

I avoided mentioning the technology itself and instead asked open-ended questions about their eating habits, grocery shopping routines, meal planning, and overall relationship with food. The goal? To map out their daily food journey and identify the pain points they experience around nutrition, convenience, and food security.

Key Takeaways:

• People adapt their diets based on those they live with—typically, the most restrictive diet wins.
• Cue to action—eating habits often stem from a single defining moment (e.g., a surgery, book, Netflix series, retreat, or health scare).
• Dietary restrictions tend to increase and become more selective over time.
• People visit multiple grocery stores—typically 3 different stores to get all the ingredients they need.
• Time spent on food: ≥1 hr / day prepping and 1-2 hrs grocery shopping / week.
• Most people view food as medicine—they believe that "eating better = feeling better."
• Ingredient-first shopping—People buy ingredients first, then craft a recipe (bottom-up approach).
• Meal planning is minimal—most people don’t plan meals more than a few hours in advance.
• Transparency matters—Consumers prefer to see exactly what’s in their food and avoid highly processed ingredients.

Ideation

Most 3D printers on the market are designed for plastics and metals and, as a result, tend to be boxy and rigid due to their Cartesian coordinate systems (X/Y/Z movements). But healthy foods aren’t geometrically perfect or rigid–they’re organic. So, I explored a cylindrical coordinate system for food printing instead.

Why?

• Better suited for organic shapes.
• Aligns with the circular geometry of plates and bowls.
• Subtly recalls the rotational turntable of a microwave oven.
• Looks less "industrial" and more like "kitchen-friendly."
• Makes printed food feel more approachable.

This shift in design philosophy played a major role in enhancing the perception of 3D-printed food and creating a more intuitive, user-friendly cooking experience.

Conceptualization

It was important to design an open glass face to keep the process transparent, building trust with the consumer as the food is being 3D-printed. This way, they could see the entire process unfold, creating a more engaging and enjoyable experience with the final-printed food product. Additionally, the system needed to integrate laser cooking, allowing for selective, in situ cooking during the 3D printing process.

These models evolved from purely aesthetic representations to fully functional concepts, incorporating motors, laser diode placements, protective shielding, and more advanced cartridge designs for improved performance.

Cartridge Design

A major pain point with typical food-grade syringes is the cleaning process—viscous pastes tend to cake along the inner barrel, making them tedious to clean by hand. To solve this, I designed a removable inner barrel to hold the food, paired with a rigid outer shell that locks into place using a clever sealing mechanism that can be tightened with a quarter. The shell and stopper were fabricated using an FDM 3D printer and designed to interface with two rubber gaskets and an off-the-shelf plastic tube from McMaster-Carr. When fully assembled and filled with peanut butter, it required significant force but remained completely leak-proof, demonstrating strong potential for a full production model. The shell is intended to slip off to be cleaned in a dishwasher or by hand and the inner cylinder would be disposable, and made from a more sustainable material.

Designing Future Foods

Food is deeply emotional and instinctual—for thousands of years, we’ve been conditioned to recognize what’s fresh, healthy, or even dangerous based on appearance, smell, and taste. But what happens when we introduce a new method of food creation—one that enables limitless designs and flavor combinations previously impossible? Do we embrace these innovations, or do they become too unfamiliar to be palatable?

Another challenge with 3D-printed foods is structural feasibility. Just because you can design something doesn’t necessarily mean that you can print it. Unlike plastics or metals, which allow for support structures to create intricate overhangs and lattices, printed foods are constrained by the shear-thinning properties of the materials themselves. The very nature of the food—its viscosity, elasticity, and flow behavior—is a significant design constraint.

We experienced this challenge firsthand with one of the most complex multi-ingredient prints to date–a seven-ingredient slice of cake. While it may sounds simple, balancing structure, flavor compatibility, aesthetic beauty, and printability in a single design proved to be an intricate process. Ensuring that the physical print matched the digital design required multiple iterations, refining the structural components of the design to achieve a final product that was both aesthetically and functionally successful.

Simulating Prints

Since dynamic viscosity is the key rheological factor affecting a food’s ability to hold its shape once printed, we saw value in developing a dynamic simulation engine to test recipe files (G-code). This tool would streamline the ideation process for multi-ingredient prints, allowing us to diagnose structural failure points and validate digital recipes—all without the need for physical printing.

We built this engine using Bifrost in Autodesk Maya, leveraging its particle-based simulation capabilities to define the dynamic viscosity of different ingredients and accurately model their printing behavior in a virtual environment.

Software Ecosystem (UI / UX)

Developing a machine without supportive software is like having an iPod without music—it just doesn’t work. For a long time, I envisioned an app to accompany this technology, but it wasn’t until Cameron, a talented software engineer, joined our team that we were able to bring it to life. Our goal was to understand how people might interact with this kind of technology daily, so we started by developing a user flow.

We aimed for a minimalist, intuitive design—one that felt simple, engaging, and easy to navigate. For inspiration, we looked at Spotify’s app, since discovering and curating meals felt similar to discovering and favoriting music. The experience needed to be fun, functional, and familiar, making meal creation as effortless as browsing a playlist.

Towards the end of this exploration, Cameron expanded the mobile app into a full-blown tablet UX. This version leaned more toward a food delivery-style experience, allowing users to build and customize their meals even when they were on the go—not just at home.

Overall, this was an incredibly fun and insightful exploration, giving us a deeper understanding of how this technology could seamlessly integrate into daily life.

Functional Design

While the broader research scope of this project meant that some concepts remained digital, we successfully built a large-scale, 18-ingredient machine capable of dual-laser cooking using off-the-shelf components. This served as a functional testbed for multi-ingredient food printing, allowing us to explore real-world applications of this technology.

To drive the system, we also developed a custom slicer engine to convert digital models into printable recipe files. Using Slic3r as a base, we customized the open-source slicer to optimize performance for our specific printer and ingredient properties. We dubbed our version Juli3nne, which is specifically tailored specifically for the 18-ingredient machine pictured above (available for use here).

With this functional setup, we designed and printed a variety of digitally-crafted foods to showcase how food printing and laser cooking can work in tandem to create intricate multi-ingredient meals.

We even worked with an NY-based chef to craft this special meal to highlight how this technology might be able to augment traditional meal creation. And the printed key lime pie (below) was designed for TF1, a French TV outlet.