1. Mapping the Invisible: The Role and Refinement of GPS Data
Modern urban planning has moved beyond guesswork. By leveraging high-resolution GPS data from shared mobility fleets, researchers can now visualize the "invisible" flows of a city.
- The Challenge of Data Quality: From a technical perspective, raw GPS data is often "noisy" due to signal interference in dense urban environments (the urban canyon effect). Noise reduction and the accurate refinement of high-quality data are the most critical steps in ensuring the reliability of any analysis.
- Trajectory Tracking: Once the data is properly cleaned, it allows us to identify not just the start and end points, but the exact paths riders take, revealing which streets are preferred and which are avoided due to poor infrastructure.
- Dwell Time Analysis: Understanding where vehicles remain stationary helps city planners optimize parking zones and prevent sidewalk clutter.
2. Temporal Dynamics: When Do People Ride?
Usage patterns are highly sensitive to the clock. Data analysis reveals 두 distinct peaks in daily micro-mobility demand.
- The Commuter Peak: A sharp increase is observed during morning ($08:00–09:00$) and evening ($18:00–19:00$) hours, confirming that e-scooters are primarily used for "first and last mile" transit.
- The Leisure Surge: On weekends, the pattern shifts toward a broader midday peak, suggesting a transition from functional commuting to recreational use.
3. Spatial Concentration: The "Hotspot" Phenomenon
Mobility is rarely distributed evenly across a city. Spatial analysis using heatmaps typically reveals high concentration in specific "hubs."
- Transit Integration: Data shows massive clusters around subway stations and major bus terminals, reinforcing the role of micro-mobility as a feeder system for public transit.
- Commercial and Campus Zones: High density is often found in business districts and university areas, where short-distance travel between buildings is frequent.
4. The Seasonality Factor: Weather and Mobility
Unlike cars, micro-mobility is deeply affected by environmental conditions.
- The Temperature Threshold: Research indicates a strong correlation between ambient temperature and ridership. Demand typically peaks in late spring and early autumn.
- Winter Decline: In colder seasons, usage can drop by as much as $50\%$ to $70\%$, necessitating different fleet management strategies for operators during the winter months.
Conclusion: Data Quality as a Foundation for Policy
Understanding these patterns is not just an academic exercise. However, the integrity of these insights depends entirely on the precision of data pre-processing. By turning "noisy" raw data into refined, actionable intelligence, we can design safer and more efficient urban environments. As a researcher, I believe that high-quality data refinement is the true key to unlocking the potential of smart cities.