Skip to main content
Industrial AI Software — Windows

SpadanaDepanelVision

AI-Powered Bridge Detection & CNC Path Generation

YOLO Vision AI + Hybrid Tangent Refinement for CNC-Grade Precision

Auto-Detect Active
SpadanaDepanelVision Interface showing automated DXF alignment and CNC path generation

CNC-Grade Precision

While YOLO Vision detects bridge locations in seconds, our proprietary Hybrid Tangent algorithm snaps them exactly to the true DXF geometry for flawless routing.

Interactive Comparison
Before AI Processing
After AI Processing with Hybrid Tangent Snapping

Drag the slider to see the CNC-grade precision of Hybrid Tangent Snapping.

Product Demonstration — 3:08 Min

Watch: DXF Import to CNC-Ready Export in Under 60 Seconds

  • Bridge detection: YOLO Vision AI automatically identifies all valid bridge locations on any PCB panel geometry.
  • Hybrid Tangent refinement: bridges snap precisely onto true DXF geometry — line and arc structures handled automatically with CNC-grade precision.
  • Export: structured JSON (including start point, end point, radius, arc direction, G-code) + operator text report generated in one click.
Built for: Process Engineers · CAM Programmers · EMS Production Manager
Seen enough? Book a live demo on your own DXF files →
Workflow

Six Steps. Under 60 Seconds.

Every step that previously required physical machine interaction is now handled computationally.

1

Import DXF

Press one button to load your PCB DXF file. No manual measurement or fixture adjustment required.

2

AI Bridge Detection

YOLO Vision AI analyzes PCB geometry and detects all valid bridge locations automatically.

3

Hybrid Tangent Refinement

Bridges are snapped precisely onto the true DXF geometry. Arc and line structures handled flawlessly.

4

Optimal Milling Path

Automatically generate the shortest CNC-ready routing path between start and end points.

5

Review & Adjust

Optional manual override. Drag bridges or adjust routing if needed.

6

Export CNC Package

Export JSON, milling path, and operator report.

Technology

Core Capabilities

YOLO Vision AI

Custom-trained neural network detects valid bridge locations on any PCB panel geometry — including irregular, dense, and organic shapes.

Hybrid Tangent Engine

Proprietary geometric algorithm refines and snaps detected bridges precisely onto the true DXF geometry, handling both line-based and arc-based structures.

DXF Layer Extraction

Automatically extracts milling contours from CAD layers. No manual coordinate input. The software reads your engineering data directly.

Structured Export Package

Every detected bridge is exported as structured JSON (with geometry parameters, arc direction, radius) plus G-code and a human-readable operator report.

100% Offline Operation

No cloud connection, no data transfer, no subscription lock-in. All AI models run locally on your production hardware.

Vision-Guided Alignment

CCD fiducial recognition aligns the digital cutting path to real-world board coordinates, accounting for panel shrinkage or expansion.

Technical Specifications

Everything your integration team needs to know.

Bridge Detection EngineYOLO Vision AI (custom-trained)
Geometry RefinementHybrid Tangent Optimization Engine
Geometry SupportLines + Arcs (all PCB contour types)
Input FormatDXF (industry-standard CAD)
Output FormatsJSON (structured) · G-code · Text Report
Universal Export ReadyZero-point based coordinate geometry (JSON & G-code), compatible with all standard CAM software.
Custom IntegrationEnterprise post-processor development available for native CNC controller syntax.
Cloud DependencyNone — 100% offline, on-premise
Operating SystemWindows
Setup Time< 60 seconds per product changeover

Standardized Output Formats
(Zero Post-Processing Required)

Directly feed into your CNC milling machine.

{
  "bridges": [
    {
      "label": "B1",
      "type": "LinePath",
      "startX": 9.2678,
      "startY": 134.8321,
      "endX": 12.8675,
      "endY": 139.1222
    },
    {
      "label": "B10",
      "type": "ArcPath",
      "startX": 171.5329,
      "startY": 186.4149,
      "endX": 176.9616,
      "endY": 186.4598,
      "I": 2.797,
      "J": -9.9655,
      "radius": 10.3506,
      "gCode": "G3"
    }
  ]
}