← Back to Articles
Mechanical & Creative Automation 2024 Private Project

CNC Painting Robot

Converted my MPCNC into a multi-color painting robot with gravity-fed tool system and 48-color palette, creating wall art with precision impossible to achieve by hand.

Project Overview

After mastering CNC milling, I wanted to push the machine into a completely different domain: art. By converting my MPCNC into a painting robot, I created a system capable of producing multi-color artwork with precision and repeatability far beyond what I could achieve manually.

The project combined mechanical design (tool-changing system), software engineering (G-code generation), and creative problem-solving (how to translate digital art into machine instructions). The resulting artwork exceeded my manual drawing capabilities and decorated my walls before I moved for my master's program.

Technical Achievement: Manual tool-changing system with self-aligning, gravity-fed pen/brush holders working across 48 colors with both pencils and acrylic paint.

Technical Design

Tool-Changing Mechanism

The heart of the system was a simple but effective tool-changing design:

  • Self-aligning holders: Drop-in tool holders that positioned each pen/brush consistently
  • Manual switching: I would swap tools between color layers (automated switching proved too complex)
  • Gravity-fed dispensing: Tools held vertically, paint/ink flowed by gravity and surface tension
  • Calibrated Z-height: Each tool programmed with precise contact pressure for consistent line weight

Color System

I worked with a palette of 48 base colors:

  • Pencils: Standard colored pencils that produced very consistent results
  • Acrylic paint: Brushes with fluid paint, more challenging but achieved painterly effects
  • Color mapping: Digital artwork color-separated into available palette colors
  • Layering strategy: Planned drawing order to minimize smudging and maximize clarity

Material Challenges

Different media presented distinct engineering problems:

  • Pencils: Easy to control, predictable, excellent for detail work
  • Acrylic paint: Inconsistent flow during large rapid movements; paper degradation over time from repeated passes
  • Paper saturation: Too much paint caused warping and bleeding
  • Drying time: Had to sequence colors to avoid wet-on-wet smearing

Technologies Used

CNC Conversion
Tool Design
G-code Generation
Prusa Slicer Hacking
Multi-color Workflows
Creative Automation

The Software Challenge: Generating G-code

The Problem

Converting artwork into CNC machine instructions proved harder than the mechanical build:

  • CNC software limitations: Standard CAM programs were painfully slow for artistic paths
  • Multi-color complexity: No existing tool designed for pen-plotter multi-color workflows
  • Tool changes: Needed coordinated pause points for manual tool swapping

The Solution: Hacking Prusa Slicer

I discovered an unconventional solution: repurpose 3D printing software for 2D art:

  1. Artwork preparation: Color-separate artwork into individual SVG files per color
  2. Import as 3D objects: Load each SVG into Prusa Slicer as if it were a 3D model
  3. Single-layer slice: Configure for one-layer-high "print" (essentially a 2D path)
  4. Multi-object slicing: Assign different "colors" to each SVG layer
  5. Automatic tool changes: Slicer generates pause commands between colors for tool swapping
Why This Worked: Prusa Slicer's multi-material workflow already solved the orchestration problem—I just applied it to pens instead of plastic extruders.

Profile Modifications

I customized Prusa Slicer settings for drawing instead of 3D printing:

  • Disabled all Z-axis movements except initial positioning
  • Adjusted "extrusion" parameters to control line weight
  • Tuned travel speeds to prevent paint splatter
  • Added custom G-code for tool-change pause points

Advanced Technique: Shading Experiments

Multi-Layer Color Mixing

I experimented with creating additional colors through layering:

  • Concept: Stack multiple pencil colors to create blended shades
  • Implementation: Slice artwork at multiple "heights" with different colors per layer
  • Results: Successfully created depth and shading effects beyond the 48 base colors
  • Limitation: Abandoned further work after achieving proof of concept

Why It Matters

The layering technique demonstrated that the system wasn't limited to flat coloring—it could produce genuinely artistic effects with proper G-code programming. This opened possibilities for:

  • Gradient effects through controlled overlap
  • Texture through varied line density
  • Depth perception through color layering

Results & Artwork

Output Quality

The finished pieces consistently exceeded my manual drawing ability:

  • Precision: Sub-millimeter positioning accuracy impossible to achieve by hand
  • Repeatability: Could reproduce identical artwork on demand
  • Detail level: Maintained fine details across large compositions
  • Time investment: Hours of machine time, but zero manual skill required during execution

Personal Impact

Several pieces decorated my apartment walls before relocating for my master's program:

  • Created artwork that I genuinely valued as decoration
  • Demonstrated that technical skill can enable creative expression
  • Proved that unconventional tool use (hacking slicer software) solves real problems
Creative Validation: These were genuinely better than anything I could draw manually—the machine amplified my artistic intent without requiring traditional artistic skill.

Key Learnings

Creative Automation

  • Tool repurposing: Machines designed for one purpose can excel at completely different tasks
  • Software creativity: The hardest part wasn't mechanics—it was translating art into machine instructions
  • Unconventional solutions: 3D printing software worked better than purpose-built CAM tools

Technical Skills

  • G-code mastery: Deep understanding of how to orchestrate complex multi-step processes
  • Software adaptation: Learned to bend existing tools to solve new problems
  • Workflow design: Created efficient pipelines from digital art to physical output

Material Science

  • Media properties: Different drawing materials have vastly different machine requirements
  • Substrate interaction: Paper quality and saturation limits matter for multi-pass work
  • Process tuning: Each material required custom speed, pressure, and sequencing parameters

Challenges & Limitations

What Worked Well

  • Pencil artwork: Highly reliable, excellent results
  • Manual tool changing: Simpler than automated systems, perfectly adequate
  • Prusa Slicer workflow: Faster than traditional CAM software by orders of magnitude

What Struggled

  • Paint consistency: Acrylic flow varied with large rapid movements
  • Paper degradation: Too many passes damaged the paper surface
  • Drying coordination: Had to carefully sequence layers to avoid smearing

Why I Stopped

The project achieved its core goal—demonstrating that creative automation was possible and producing artwork I valued. I moved on to other projects rather than refining the system further, but the techniques remain documented for future use.

Impact & Applications

This project influenced my approach to several subsequent builds:

  • Tool versatility: Machines should be designed for adaptation and repurposing
  • Software creativity: Often the best solution is repurposing existing software for new domains
  • Creative engineering: Technical skills can enable artistic expression without traditional training
  • Workflow automation: The principles apply to any multi-step creative or manufacturing process

Related Projects

This build connects to several other mechanical and creative works: