I Tested Illustrae: Can AI Actually Design Scientific Figures?

By Andy Stapleton

Share:

Key Concepts

  • Illustring: An AI-powered design platform specifically tailored for academic needs, including graphical abstracts, research posters, and scientific figures.
  • Canvas-based Workflow: A large, flexible workspace that allows for the integration of multiple AI-generated figures, text, and external data.
  • Brainstorming Tool: An automated feature that generates poster or abstract layouts based on pasted research text.
  • Element Generation: The ability to create specific scientific illustrations (e.g., organic photovoltaic devices, microscopic structures) via text prompts.
  • Raster vs. Vector: The distinction in export formats, where PNG is used for AI-generated raster images and SVG is used for scalable shapes and text.

1. Platform Overview and Interface

Illustring addresses a significant gap in academic AI tools by focusing on visual communication. The interface is organized into:

  • Side Panel: Access to canvases, templates, tutorials, and a library of pre-made scientific assets.
  • Central Workspace: A massive canvas that supports multiple figures, allowing users to build complex, multi-part illustrations.
  • Brainstorming Module: Users input their research title and abstract to receive a structured, "boxy" layout for posters or graphical abstracts. The tool allows for customization of style (e.g., modern, realistic, watercolor) and color schemes, including the ability to input institutional brand colors.

2. Methodology: Creating Scientific Figures

The platform allows for two primary workflows: automated brainstorming or manual construction from scratch.

Step-by-Step Process for Custom Figures:

  1. Define Components: Break down the research process into individual elements (e.g., a specific device layer or a process step).
  2. Prompting: Use the top input bar to describe the element. The AI demonstrates a "scientific perspective," often correctly identifying technical components (e.g., layers in an organic photovoltaic device like P3HT, PCBM, and electrodes).
  3. Style Selection: Choose from predefined styles or upload a reference image to guide the AI’s output.
  4. Editing: Use the "Edit" function to refine specific parts of an image. Note: The tool does not maintain a history, so users must restart from the base element if an edit is unsatisfactory.
  5. Refinement: Use standard image processing actions (crop, flip, bring to front/back) to arrange elements.
  6. Combination: Select multiple elements and use the prompt bar to "combine" them into a cohesive, multi-step illustration.

3. Advanced Features and Real-World Applications

  • Poster Design: The tool can generate full-scale academic posters. It demonstrates an understanding of academic layout conventions, including the placement of SEM (Scanning Electron Microscope) imagery and data visualizations.
  • Library Integration: Users can pull from a library of pre-existing biological, medical, and equipment-related icons.
  • External Data: Users can upload their own data figures or photos to incorporate into the canvas, allowing for a hybrid of AI-generated art and real research data.
  • Frame Tool: By pressing 'F', users can group elements into a frame, which simplifies the export process for specific sections of a large canvas.

4. Technical Specifications and Exporting

  • Exporting: Users can export as PNG (for AI-generated graphics) or SVG (for shapes and text).
  • Export Options: The platform supports "Export Only Selected," "Remove Background," and "Dark Mode" for posters.
  • Pricing: The service offers a tiered subscription model, starting at approximately $8.99 AUD/month, which includes 50 image generations.

5. Notable Quotes

  • "It’s scarily good because that is exactly what they look like." — Regarding the AI's ability to generate accurate representations of silver nanowires under a microscope.
  • "If you are able to describe the image, you are able to produce a figure." — Highlighting the accessibility of the prompt-based generation system.

6. Synthesis and Conclusion

Illustring serves as a powerful, specialized tool for researchers who struggle with the design aspects of academic dissemination. Its ability to understand scientific context—such as the specific layers of a solar cell or the structure of a graphical abstract—sets it apart from general-purpose image generators. While the platform has minor limitations (such as the lack of an "undo" history and occasional issues with mobile photo uploads), its capacity to streamline the creation of professional-grade posters and figures makes it a highly actionable asset for academic workflows.

Chat with this Video

AI-Powered

Hi! I can answer questions about this video "I Tested Illustrae: Can AI Actually Design Scientific Figures?". What would you like to know?

Chat is based on the transcript of this video and may not be 100% accurate.

Related Videos

Ready to summarize another video?

Summarize YouTube Video