Embroidery Grammar : Path Decision System v1

Overview

The Path Decision System v1 is a rule-based computational framework designed to simulate embroidery path selection through structured logic.
It transforms embroidery from a manual craft process into a decision-driven system, enabling AI to interpret and generate stitch sequences based on defined rules.

Core Concept

Embroidery is understood as acomputable structural language system，It consists of the following three elements:
✦Nodes：entry/exit points
✦Paths：the sequence of thread movement
✦Tension：stability and constraint conditions
The system evaluates possible next steps at each node and selects the optimal path based on a scoring mechanism.。

System Structure

At each step, the system processes:
✦Current Node
✦Prev Node(memory)
✦Available Options（Option1, Option2）
Each option is evaluated using a multi-factor scoring model.

Scoring Mechanism

Each candidate path is assigned a score based on the following components:
✦Structure Score:Validity of the path within the defined node network
✦Tension Penalty:Penalizes immediate backtracking to the previous node
✦Continuity:Rewards smooth and forward-moving paths
✦Backtrack Penalty:Strongly discourages reversal behavior

Decision Rule

The system selects the path with the higher total score:
✦If Score1 > Score2 → choose Option1
✦If Score2 > Score1 → choose Option2
If equal → apply tie-breaking strategy (e.g., alternate or random selection)

System Behavior

With memory and scoring applied, the system demonstrates:
✦Avoidance of immediate backtracking
✦Continuous path formation
✦Emergence of cyclic structures (pattern loops)
✦Directional path consistency

These behaviors correspond to fundamental embroidery principles such as：
✦thread continuity
✦structural stability
✦repeatable patterns

Theoretical Significance

This system represents a shift from:
Embroidery as technique→ toEmbroidery as a computable language

It establishes a foundational model for:
✦AI-based embroidery analysis
✦Generative pattern systems
✦Structural grammar in textile logic

Future Development

Planned extensions include:
✦Multi-step memory (path history)
✦Center bias and spatial awareness
✦Closure detection (completion logic)
✦Pattern recognition and repetition modeling