Solar Roof AI Planner

An interactive solar planning workspace for rooftop mapping, AI-assisted roof detection, panel layout simulation, and financial estimation.

---

💡 Why This Project Exists

Solar feasibility tools are often fragmented across GIS software, internal spreadsheets, and engineering workflows. This project demonstrates how a single app can guide users from location discovery to actionable planning output, while keeping AI suggestions reviewable and editable.

This is not just a detection demo. It's a complete product-style workflow that balances:

• ✨ Interactive map UX
• 🤖 Assistive computer vision
• 📐 Geospatial calculations
• 💰 Planning and financial context

---

📸 See It In Action

Workspace Overview

Blueprint

PDF Report

---

⚡ Core Features

1️⃣ Property Search & Mapping

• Address & Place Search: Fast rooftop lookup.
• Map-first Workspace: Centered automatically on the selected property.
• Satellite Analysis Flow: Visual roof clarity with high-res imagery.

2️⃣ Roof & Obstacle Mapping Tools

• Manual Drawing: Interactive map editing tools.
• Separate Layers: Editable layers for roof polygons and obstacles.
• Complex Geometries: Support for irregular shapes and exclusion zones.

3️⃣ AI-Assisted Roof Detection

• Direct Integration: Send snapshots directly to Roboflow hosted workflows for analysis.
• Plane & Obstacle Detection: Identify roof structures directly from imagery.
• Review & Apply: Preview detections before applying to the workspace. Tune confidence and area filters.

4️⃣ Geospatial & Solar Analysis

• Metrics Computation: Calculate gross area, blocked area, and net usable area.
• Orientation Estimation: AI-driven orientation candidates.
• Solar Exposure Heatmap: Overlay visual zones showing stronger and weaker exposure areas based on sun position and context.

5️⃣ Panel Layout Simulation

• Manual & Auto-pack Placement: Fine-grained manual control or automatic maximization.
• Capacity-Aware Selection: Select panel types with real-time capacity counting.
• Validation: Prevent overlaps and invalid placements using worker-based operations for responsive UI.

6️⃣ Financial Planning & Export

• Financial Dashboard: Sizing inputs, planning assumptions, and real-time estimation charts.
• Export Options: GeoJSON export of roof/obstacle geometry.
• Reporting: Generate blueprint-style PDF reports for planning handoffs.

---

🎯 End-to-End User Flow

1. 🔍 Search for a property.
2. 🗺️ Enter the map workspace and switch to imagery mode.
3. 🏗️ Draw roof and obstacles manually, or run auto-detection.
4. ✅ Review detection results and accept only what looks correct.
5. 📊 Calculate usable roof area and inspect solar heatmap hints.
6. 📦 Simulate panel placement manually or with auto-pack.
7. 💵 Review estimated system capacity and financial outcomes.
8. 📤 Export geometry/report artifacts.

---

🏗️ Architecture

Frontend (React 18 + Vite + TypeScript)

• 🗺️ Interactive mapping and draw/edit UX (Leaflet, Turf.js)
• 🔎 Address search integration (OpenStreetMap Nominatim)
• ⚙️ Panel layout logic and worker offloading
• ☀️ Solar heatmap visualization
• 💰 Financial dashboard components (Recharts)

Hosted Detection (Roboflow Workflow)

• ☁️ Serverless workflow called directly from the browser (POST https://serverless.roboflow.com/<workspace>/workflows/<workflow_id>)
• 🧾 Structured output parsing from svg_output and json_output
• 📋 Metadata shaping into app-compatible roof/obstacle results

---

🚀 Getting Started

Prerequisites

• Node.js 20+

Installation & Execution

1. Clone the repository:

   git clone https://github.com/anomalyco/Solar-Roof-AI.git
   cd Solar-Roof-AI

2. Install dependencies:

   npm install --workspace frontend

3. Set up Environment Variables: Create a .env file in the frontend directory with the following variables:

4. Run the development server:

   npm run frontend:dev

   The app will be available at http://localhost:5173.

---

⚠️ Limitations and Assumptions

• 🖼️ Detection quality heavily depends on imagery quality, zoom level, and roof contrast.
• 📐 Pitch, aspect, and height values are estimation-grade derived from 2D imagery.
• 🎯 Results are intended for planning and pre-sales exploration, not permit-ready engineering.
• ✏️ Manual edits remain essential for complex edge cases.

---

🔮 Roadmap

• Stronger model-based detection beyond classical CV heuristics.
• More explicit setback and code-rule constraints for panel placement.
• Time-series irradiance simulation and seasonal production profiles.
• Authentication, saved projects, and collaboration workflows.

---

🤝 Contributing

Contributions, issues, and feature requests are welcome! Feel free to check issues page if you want to contribute.

---

📄 License

This project is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) License.