How It All Connects

# openclaw_tools/scan_room.py
from openclaw import Tool, Parameter

class ScanRoomTool(Tool):
    name = "scan_room"
    description = "Trigger drone to scan a room"
    parameters = [
        Parameter("room", str, "Room to scan"),
        Parameter("purpose", str, "Purpose of scan"),
    ]

    def execute(self, room: str, purpose: str):
        # 1. Validate drone status
        drone = DroneController.get_status()
        if drone.battery < 30:
            return "Battery too low for scan"

        # 2. Select waypoint mission
        mission = WaypointPlanner.plan(
            room_type=room,
            altitude=1.5,
            overlap=0.7
        )

        # 3. Execute flight
        drone.execute_mission(mission)

        # 4. Process images
        point_cloud = Photogrammetry.process(
            images=drone.captured_images,
            engine="colmap"
        )

        # 5. Return viewer URL
        url = PointCloudServer.upload(point_cloud)
        return f"3D model ready: {url}"

# telegram_bridge.py
import telebot
from openclaw import OpenClawClient

bot = telebot.TeleBot(TELEGRAM_TOKEN)
claw = OpenClawClient(api_key=OPENCLAW_KEY)

@bot.message_handler(func=lambda m: True)
def handle_message(message):
    # Forward to OpenClaw with ChatGPT
    response = claw.chat(
        message=message.text,
        model="codex-5.1",
        tools=["scan_room", "drone_status",
               "export_model"]
    )
    bot.reply_to(message, response.text)

    # If a 3D model was generated, send link
    if response.has_attachment:
        bot.send_message(
            message.chat.id,
            f"🏠 3D Model: {response.attachment_url}"
        )

#!/bin/bash
# process_scan.sh - COLMAP pipeline

INPUT_DIR="./captured_images"
OUTPUT_DIR="./3d_output"

# Step 1: Feature extraction
colmap feature_extractor \
  --database_path $OUTPUT_DIR/db.db \
  --image_path $INPUT_DIR \
  --ImageReader.single_camera 1

# Step 2: Feature matching
colmap exhaustive_matcher \
  --database_path $OUTPUT_DIR/db.db

# Step 3: Sparse reconstruction
colmap mapper \
  --database_path $OUTPUT_DIR/db.db \
  --image_path $INPUT_DIR \
  --output_path $OUTPUT_DIR/sparse

# Step 4: Dense reconstruction
colmap image_undistorter \
  --image_path $INPUT_DIR \
  --input_path $OUTPUT_DIR/sparse/0 \
  --output_path $OUTPUT_DIR/dense

colmap patch_match_stereo \
  --workspace_path $OUTPUT_DIR/dense

# Step 5: Point cloud fusion
colmap stereo_fusion \
  --workspace_path $OUTPUT_DIR/dense \
  --output_path $OUTPUT_DIR/fused.ply

echo "Point cloud: $OUTPUT_DIR/fused.ply"