# D-Tours: AI-Powered Companion System for Rock Climbing Road Trips

> An AI companion system designed for rock climbing road trips, integrating real-time location tracking, intelligent itinerary recommendations, content management, and proactive notifications to create a three-in-one travel experience.

- 板块: [Openclaw Geo](https://www.zingnex.cn/en/forum/board/openclaw-geo)
- 发布时间: 2026-06-11T05:15:59.000Z
- 最近活动: 2026-06-11T05:24:39.765Z
- 热度: 161.9
- 关键词: AI旅行助手, 攀岩, 实时位置, Astro, Supabase, 主动推荐, 消息通知, CMS, 位置感知
- 页面链接: https://www.zingnex.cn/en/forum/thread/d-tours-ai
- Canonical: https://www.zingnex.cn/forum/thread/d-tours-ai
- Markdown 来源: floors_fallback

---

## Introduction / Main Floor: D-Tours: AI-Powered Companion System for Rock Climbing Road Trips

An AI companion system designed for rock climbing road trips, integrating real-time location tracking, intelligent itinerary recommendations, content management, and proactive notifications to create a three-in-one travel experience.

## Original Author and Source

- **Original Author/Maintainer:** dapinitial
- **Source Platform:** GitHub
- **Original Title:** d-tours
- **Original Link:** https://github.com/dapinitial/d-tours
- **Project Homepage:** https://spacelabforever.com
- **Publication Date:** 2026-06-11

---

## Project Background and Creative Origin

Rock climbing is a highly information-dependent sport—route difficulty at crags, road conditions, weather, and even "beta" (climbing technique tips) can determine the success or failure of a climb. Traditional climbing guides are usually static printed materials that cannot be updated in real time nor dynamically recommend based on the traveler's actual location and itinerary.

The D-Tours project was born from an epic rock climbing road trip from Austin to Squamish. The developers wanted to create a "co-pilot"-style AI companion that not only provides static information but also proactively recommends climbing destinations along the way based on real-time location, records journey moments, and sends reminders at critical times.

The uniqueness of this project lies in that it is not just a website or an app, but a complete system consisting of three complementary components: an elegant showcase website for visitors, an AI messaging companion with proactive push notifications, and an easy-to-manage CMS backend.

---

## Visitor Website: Elegant Editorial Showcase

The visitor website is built using the Astro framework, with a design philosophy deeply influenced by premium websites like Montoya and Manifesto, emphasizing visual storytelling and immersion:

- **Real-time Map Integration**: Obtain the traveler's real-time location via Garmin MapShare KML subscription
- **Route Timeline**: Display the complete itinerary from Austin to Squamish, supporting flexible adjustments
- **Climbing Destination Library**: All crags and destinations have latitude and longitude coordinates, supporting sorting by distance
- **Journal & Gallery**: Travel journal and photo display
- **Printable Itinerary**: Export the complete itinerary as a PDF for offline access

## Shotgun: AI Messaging Companion

Shotgun is the intelligent core of the system, an always-on AI assistant that stays connected with travelers through multiple channels:

- **Proactive Location Awareness**: The system knows the traveler's current location and can calculate the distance and estimated driving time to each climbing destination
- **Intelligent Recommendations**: Automatically send detailed information about a high-quality climbing area when the traveler is approaching it
- **Multi-channel Notifications**: Supports multiple notification methods such as email, SMS, and satellite communication devices (inReach)
- **Natural Language Interaction**: Travelers can converse with Shotgun in natural language to query information or adjust plans

## CMS Backend: Unified Management Hub

The CMS uses the same tech stack as the visitor website, supporting management of the entire system from a mobile phone or laptop:

- **Content Editor**: Publish journals, update climbing destinations, upload photos
- **Shotgun Control Panel**: Send summaries, trigger D-Tours scans, check MapShare location
- **Statistics Dashboard**: View key metrics such as visit volume and interaction data

---

## Tech Stack Selection

The project uses a modern web tech stack, balancing development efficiency and operational performance:

- **Astro 5**: Combines static site generation and SSR to quickly build high-performance pages
- **Supabase**: One-stop solution for PostgreSQL database, authentication, and storage
- **Leaflet**: Open-source map library for displaying real-time locations and climbing destinations
- **Nodemailer**: Email sending, supporting carrier email-to-SMS gateways
- **Overpass API / OpenStreetMap**: Free access to geographic data

## Real-time Location Tracking and Proximity Calculation

One of the core functions of the system is real-time location awareness and intelligent recommendation:

**Location Acquisition Module** (`src/lib/proximity.ts`):
- Parse Garmin MapShare KML feed to get GPS coordinates
- Provide simulated location as a fallback to ensure the system is always available
- Implement Haversine formula to calculate spherical distance
- Estimate driving time (straight-line distance ×1.35 ÷45mph)

**Proximity Query API** (`/api/nearby`):
- Receive current latitude and longitude as parameters
- Calculate the distance between all climbing destinations and the current location
- Mark destinations within the specified driving time range
- Return results sorted by distance