Modern public transportation is no longer run on guesswork. Behind every bus schedule, route decision, and real-time arrival notification lies a sophisticated layer of data collection and intelligent planning. As cities grow and passenger expectations rise, transit agencies are turning to Intelligent Transportation Systems (ITS) to move from static surveys to dynamic, data-driven decision-making. This shift is transforming how public transit networks are designed, operated, and managed at every level.
Table of Contents
- The four stages of intelligent transport planning
- Strategic planning: long-term network design
- Tactical planning: frequencies and timetables
- Operational planning: vehicle and driver scheduling
- Real-time applications: daily control strategies
- Leveraging ITS data for tactical and operational planning
- Automated Vehicle Location systems
- Automated Fare Collection and demand patterns
- The vehicle scheduling problem enhanced by data
- Data sources for intelligent transportation planning
- GPS and GIS technologies
- Sensors and traffic detectors
- Smart card data applications
- Mobile phone and Call Detail Records
- Moving toward truly data-driven transit
The four stages of intelligent transport planning
Intelligent planning in public transportation operates through a structured hierarchy of decision-making. Researchers Iliopoulou and Kepaptsoglou (2019) outline four distinct stages that organize the planning process, each with different time horizons and objectives. Understanding these stages is essential for appreciating how ITS data flows through the entire transit ecosystem.
Strategic planning: long-term network design
At the highest level, strategic planning addresses fundamental questions about where transit lines should exist. This involves decisions about network topology, terminal locations, and corridor selection that will shape a city’s mobility for decades. These choices require understanding broad population distribution, land use patterns, and long-term urban development goals. While traditionally based on census data and comprehensive surveys, ITS data is increasingly supplementing these sources by providing actual travel behavior patterns at unprecedented scale.
Tactical planning: frequencies and timetables
Once the network structure is established, tactical planning determines how often buses or trains should run and when exactly they should depart. This stage creates the timetables and service frequencies that passengers rely on daily. Tactical decisions must balance passenger demand with operational costs, ensuring vehicles aren’t overcrowded during peak hours while avoiding wasteful empty runs during quiet periods. The goal is to provide adequate capacity throughout the day while maintaining reasonable headways between services.
Operational planning: vehicle and driver scheduling
With timetables in place, operational planning assigns specific vehicles and drivers to execute the planned service. This is where the Vehicle Scheduling Problem (VSP) comes into play-a complex optimization challenge that has occupied transportation researchers for decades. The VSP involves creating sequences of trips that individual vehicles will perform throughout the day, from leaving the depot in the morning to returning at night. Public transit systems rely on efficient scheduling to keep operating costs in check while ensuring all timetabled trips are covered.
Real-time applications: daily control strategies
Even the best plans encounter reality. Real-time applications deal with the inevitable disruptions, delays, and variations that occur during actual operations. This includes adjusting dispatching, managing bunching when buses cluster together, and responding to incidents or unexpected demand surges. Modern ITS enables operators to monitor the entire network continuously and make immediate adjustments to maintain service quality.
Leveraging ITS data for tactical and operational planning
The true power of ITS emerges when its data streams feed directly into planning decisions. Two key technologies-Automated Vehicle Location (AVL) and Automated Fare Collection (AFC)-provide the raw material for this transformation.
Automated Vehicle Location systems
AVL technology uses GPS to track transit vehicles in real-time, creating a continuous record of where every bus or train is at any moment. AVL data collected over time enables planners to identify bottlenecks, measure actual travel times between stops, and understand how running times vary throughout the day and across seasons. Before AVL implementation, transit supervisors often had to guess driver locations unless they contacted them by radio. Now, dispatchers can monitor the status and location of all equipped vehicles on their computers.
For operational planning, AVL data reveals real trip time variability-essential information for creating robust vehicle schedules. Traditional approaches assumed fixed travel times between stops, but actual running times fluctuate significantly based on traffic, passenger volumes, and other factors. By incorporating this variability, agencies can build more reliable schedules that maintain connections and reduce delays.
Automated Fare Collection and demand patterns
AFC systems using smart cards record every passenger transaction-when they boarded, where they tapped their card, and for integrated systems, where they exited. Integrating AVL with AFC data allowed New York City Transit to develop a citywide bus ridership model that analyzes trip passenger loads and full origin-destination data at both route and neighborhood levels. This represented a transformation from analysis techniques that relied primarily on professional judgment to more sophisticated statistical methods.
At the tactical level, longitudinal AFC data helps capture demand variability and traveler responses to service changes. Planners can see how ridership shifts when frequencies are adjusted or routes modified, enabling evidence-based timetable optimization. This creates a feedback loop where service changes can be evaluated objectively rather than relying solely on passenger complaints or intuition.
The vehicle scheduling problem enhanced by data
The Vehicle Scheduling Problem is a combinatorial optimization challenge that assigns trips to vehicles while minimizing operational costs. The basic objective is to cover all timetabled trips with the fewest vehicles and shortest deadhead (non-revenue) distance. Traditional approaches used deterministic models with fixed travel times between stops.
However, AVL data enables a more sophisticated approach by incorporating real trip time variability. Schedules can be designed with appropriate buffer times based on actual historical performance rather than arbitrary padding. Research is exploring how to balance operational cost minimization with schedule robustness, recognizing that tight schedules vulnerable to delay propagation can ultimately cost more through service disruptions and passenger dissatisfaction.
Data sources for intelligent transportation planning
Multiple data sources now contribute to intelligent transportation planning, each with distinct advantages and limitations. Choosing the right combination depends on specific planning needs, available budgets, and local infrastructure.
GPS and GIS technologies
Geographic Information Systems combined with GPS provide detailed spatial information about travel behavior. These technologies enable precise route tracking and spatial analysis of transit coverage. However, deploying comprehensive GPS tracking can be costly, particularly for agencies looking to gather data on general population travel patterns beyond their own vehicles. The precision is excellent, but scalability remains a challenge.
Sensors and traffic detectors
Road sensors and traffic detectors offer high-accuracy data about traffic flow, vehicle counts, and speeds. These fixed installations provide consistent, reliable measurements at specific locations. The main limitation is high installation and maintenance costs, meaning coverage is typically limited to major corridors rather than comprehensive network monitoring.
Smart card data applications
Smart cards have emerged as an invaluable source for analyzing mobility patterns in public transit. The data reveals travel patterns, transfer behaviors, and boarding frequencies across the entire network. Researchers use smart card data to estimate origin-destination matrices, analyze vehicle congestion, and understand how passengers actually use the system.
A key advantage is sample size-smart card data captures millions of transactions rather than the thousands typically reached through household surveys. However, smart card data is mode-specific, capturing only journeys on transit systems using that payment method. Users paying by cash, using mobile tickets, or traveling by other modes remain invisible in the dataset.
Mobile phone and Call Detail Records
Mobile phone data, particularly Call Detail Records (CDRs), offers extensive and relatively cost-effective location information. Every time someone makes a call or sends a text, the nearest cell tower is recorded, creating a trace of population movement across cities. CDR data can cover millions of people-the World Bank worked with data covering about 25 percent of Rio de Janeiro’s metropolitan population, compared to traditional surveys reaching only 10,000 to 30,000 households.
However, CDRs present significant challenges. The data has unequal temporal distribution related to phone usage frequency-location is only captured when users make calls, not continuously. Spatial resolution is limited to cell tower coverage areas, which can range from 200 to 10,000 meters depending on network density. A single SIM card may be shared by multiple people, or one person may possess multiple devices, complicating individual trajectory analysis.
Privacy concerns loom large over CDR applications. Research has shown that just four random location points can uniquely identify 95 percent of individuals in a database, even when data is supposedly anonymized. This creates tension between the analytical value of detailed location data and the ethical imperative to protect personal privacy. Resolving these concerns through proper anonymization and aggregation techniques remains critical before CDR use can become mainstream in transportation planning.
Moving toward truly data-driven transit
The integration of ITS data into public transit planning represents a fundamental shift from static, survey-based approaches to dynamic, continuous optimization. Rather than planning based on occasional snapshots of travel behavior, agencies can now observe and respond to actual patterns in near real-time. This creates opportunities for more responsive service that adapts to changing demand patterns, whether seasonal fluctuations, special events, or long-term demographic shifts.
Yet challenges remain. No single data source provides complete information, and integrating multiple sources requires sophisticated analytical capabilities. Privacy protections must be robust without undermining analytical utility. And ultimately, data must translate into decisions-requiring organizational capacity to act on insights generated through ITS applications.
What do you think? As transit systems become increasingly dependent on passenger data for planning and optimization, how should agencies balance the operational benefits of detailed tracking against individual privacy concerns? And in your city, do you see evidence that transit planning has become more responsive to actual travel patterns?
References
- https://www.iso.org/transport/its-intelligent-transportation-systems
- https://www.hindawi.com/journals/jat/2019/4212631/
- https://www.transitwiki.org/TransitWiki/index.php/Automatic_vehicle_location
- https://www.sciencedirect.com/science/article/pii/S1077291X22000996
- https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9231760/
- https://blogs.worldbank.org/en/transport/big-data-comes-transport-planning-how-your-mobile-phone-helps-plan-rail-line
- https://www.flowgeek.org/what-are-cdrs/what-are-the-strengths-and-limitations-of-cdr-data
Leave a Reply