Roadkill: Nesting Laser-Cut Objects for Fast Assembly

Title of the Paper

Roadkill: Nesting Laser-Cut Objects for Fast Assembly

Paper Information

• Research Area: Human-Computer Interaction (HCI), rapid manual assembly, laser cutting, and personal fabrication
• Keywords: rapid prototyping, laser cutting, manual assembly, fabrication techniques, planning algorithms, subassemblies, CAD software, human-computer interaction, design automation, algorithm optimization

Research Background and Problem Statement

• Laser cutting offers significant advantages in rapid prototyping, especially when compared to 3D printing, as the design and cutting processes are much faster. However, the assembly process becomes a bottleneck in the overall workflow. For instance, in experiments with a simple chair model, drawing and cutting took only 3 minutes, while assembly required an average of 22 minutes, accounting for 80% of the total production time.
• Existing tools (e.g., FlatFitFab and Kyub) only provide traditional sheet layouts with manual numbering for each pair of edges. This approach relies on users visually searching and matching numbers, which becomes increasingly complex as the number of sheets increases, resulting in quadratic growth in search difficulty and severely limiting assembly efficiency.
• To further accelerate the application of laser cutting, addressing the inefficiency of the assembly process is urgently needed.

Solution

• Core Method:
  • Roadkill is a software tool that improves assembly efficiency by generating optimized 2D cutting layouts. Specifically, Roadkill embeds assembly information and guidance directly into the cutting layout using a visual language.
  • The core idea is to eliminate the efficiency bottleneck by reducing the time required for visual searching.
• Innovations:
  1. Introduced a visual language comprising 10 design elements (e.g., graphical thumbnails, numbered sequences, directional arrows, exposed connection points) to intuitively convey assembly steps.
  2. Utilized algorithms to decompose 3D models into "subassemblies." These subassemblies can be assembled independently and then gradually integrated into the complete model.
  3. Designed "Break-away Tabs" to secure subassemblies and simplify their separation.
  4. Effectively arranged subassembly sheets through direct adjacency or arrow connections, avoiding complex searches.
• Implementation Steps and Key Techniques:
  1. Algorithm Workflow:
     • Parse the 3D model to generate a "Disassembly Tree," determining reversible assembly sequences.
     • Segment complex models into independently assemblable subcomponents.
     • Optimize sheet layouts to maximize direct adjacency or arrow connections between sheets.
     • Construct complete SVG cutting layouts containing visual language information.
  2. Subassembly Generation:
     • Analyze constraints between sheets to automatically generate subassembly units.
     • Add assembly sequence labels to subassemblies and use arrows to indicate assembly directions.
  3. Design Elements Supporting Rapid Assembly:
     • Preview thumbnails provide users with the location of the subassembly within the complete model.
     • T-slot symbols and asymmetric joint designs reduce user errors during insertion and removal.

Research Outcomes

• Experimental Results:
  • Compared to traditional assembly methods, layouts generated by the Roadkill software enabled users to assemble models 2.4 times faster.
  • Specifically:
    • Assembling a chair model using traditional methods took 22 minutes, whereas Roadkill layouts reduced the average time to just 10 minutes.
    • Participants included both experienced and inexperienced users, and results showed significant improvements for all users.
  • Roadkill was tested on popular models from the Kyub library, achieving a success rate of 92% in generating 2D layouts. The average computation time was only 0.89 seconds, with complex models taking approximately 2.81 seconds.
• Subjective Feedback:
  • Users generally found Roadkill-generated layouts intuitive and straightforward. Thumbnails, arrows, and molecular assembly designs effectively conveyed assembly sequences.
  • Some participants reported minor challenges with multi-component assembly, but overall, the experience was superior to traditional methods.
• Limitations:
  1. Roadkill cannot handle models with intersecting connections.
  2. For highly interlocked structures, the number of subassemblies may become excessive.
  3. To provide annotation information, Roadkill layouts consume 37% more material compared to traditional methods.
• Future Directions:
  1. Explore combining laser cutting with origami techniques for model design.
  2. Optimize existing algorithms to reduce material waste and enhance support for complex structures.

Conclusion

Roadkill significantly enhances the efficiency of manual assembly in laser cutting through innovative sheet layouts and clear visual assembly language. Its "divide-and-conquer" approach and visual guidance demonstrate the potential for workflow optimization in personal fabrication.