Dataset of Idling Positions and Derived Trips of Shared Mobility Vehicles in Munich, Germany (2023–2025)

Description

This dataset provides high-resolution spatio-temporal data on shared mobility vehicles in Munich, Germany, collected between June 1, 2023, and May 31, 2025. It includes:

• Idling data, identifying stationary periods of vehicles based on spatial clustering of consecutive GPS positions.
• Trip-level data, representing movements between idling locations, filtered by distance and duration.

The dataset covers five providers across three shared mobility modes:

• Car-sharing: Miles, ShareNow
• Bike-sharing: MVG Rad  
• E-scooter-sharing: TIER, VOI

Geographic Scope

• City: Munich, Germany  
• Latitude: 47.9° N to 48.4° N  
• Longitude: 11.15° E to 11.9° E  

Files Included

1. idling_{provider}.parquet.gz

Content: Idling periods derived from position data  
Criteria: Stationary within 100 m radius  
Columns:
- id: Vehicle ID  
- lat: latitude (EPSG:4326) of the vehicle’s idling location
- lon: longitude (EPSG:4326) of the vehicle’s idling location
- starttime: unix timestamp of the vehicle’s idling start time
- endtime: unix timestamp of the vehicle’s idling end time

2. trips_{provider}.parquet.gz  

Content: Derived trips between idling periods  
Criteria: Distance >= 100 m, duration <= 6 hours  
Columns:
- id: Vehicle ID  
- startlat: latitude (EPSG:4326) of the trip's departure position
- startlon: longitude (EPSG:4326) of the trip's departure position
- starttime: unix timestamp of departure
- endlat: latitude (EPSG:4326) of the trip's arrival position
- endlon: longitude (EPSG:4326) of the trip's arrival position
- endtime: unix timestamp of arrival

3. vehicles_{provider}.parquet.gz  

Content: Vehicle-specific information
Columns:
- id: vehicle ID
- vehicle_type: vehicle’s model specification
- fuel_type: vehicle's primary energy source
- color: vehicle color
- time_first_seen: unix timestamp of vehicle’s first appearance in the data
- time_last_seen: unix timestamp of vehicle’s last appearance in the data

4. service_area_{provider}.parquet.gz  

Content: Service Areas 
Columns:
- provider
- geom_service_area: multipolygon of provider’s service area (EPSG:4326)

5. scraped_urls.json

Content: List of all queried URLs

Key Statistics

Data Collection & Processing

- Source: move.mvg.de (now offline)  
- Method: Python-based web scraping in 3-minute intervals  
- Coverage: Munich area divided into overlapping grid cells  
- Storage: PostgreSQL database  
- Idling detection: Based on spatial clustering within 100 m  
- Trip detection: Transitions between idling periods, filtered by distance and duration  

Validation

Trip data for MVG Rad was partially validated against official open data.  
In June 2023, 88.2% of official trips were matched with derived trips. 

Limitations

- Gaps may occur due to scraping interruptions, reservations, or round-trips.
- Some ShareNow vehicle IDs show unusually short trip durations.
- No raw position data, route, pricing, or user data is included.