Improving public transport and rail design, operations and service by big data

Improving

public transport and rail

design, operations and service

by Big Data

Integrating Passenger Information, London 2014

Dr. Niels van Oort

Assistant professor Public Transport

Dutch approach

• Using Big Data to improve public transport

• Direct approach: privacy issues tackled

• Reducing costs

• International orientation

• Paradigm shifts– Passenger -> Traveller

– Trip -> Journey

– Evaluation -> Prediction

– Two practical applications

Developments in public transport and rail industry• Focus on cost efficiency• Customer focus• Enhanced quality

Main challenges:Increasing cost efficiencyIncreasing customer experienceMotivating new strategic investments

What do we want to know?

What do you actually want to know?

Where people are– When, at what moment– How much– How long are they there– How did they get there– Where did they come from– Where are they goingand– Why are they on the move?

Understanding peoples behaviourandPredicting future scenarios

Big Data

GSM data; tracking travellers

- Potential public transport services

Crowd data

- Evaluating and optimizing pedestrian flows

Passenger data (APC); tracking passengers

- Evaluating and optimizing ridership and passengers flows

Combining data sources

- Passenger experience

Big Data

The potential benefits

Optimizing network, timetable design and services:

The Netherlands: Potential cost savings: > €50 million

Utrecht: € 400.000 less yearly operational costs

The Hague: 5-15% increased ridership

Amsterdam: ~10% increased cost coverage

Tram Maastricht:> €4 Million /year social benefits

Tram Utrecht: : €200 Million social benefits

DataData InformationInformation KnowledgeKnowledge

ImprovementsImprovements

The challenge

EvaluationEvaluation ForecastsForecasts

- Monitoring and predicting passenger numbers: Whatif- Monitoring and analysing traveller behaviour

Applications

Passenger flows and whatif

Passenger data

Connecting to transport model:• Evaluating history• Predicting the future• Elasticity approach (quick and low cost)

• Whatif scenario’s– Stops: removing or adding– Faster and higher frequencies– Route changes

• Quick insights into– Expected cost coverage– Expected ridership

Occupancy train line Zwolle-Emmen

fictitious data

Whatif scenarios

Adjusting

- Speed

- Fares

- Time of operations

- Number of stops

- Routes

- Frequency

Illustrating impacts on (indicators):- Cost coverage- Occupancy- Ridership- On time performance- Revenues- Fuel consumption

Whatif: changing the schedule

Results: Flows rerouting

fictitious data

Results: Flows increased frequencies

fictitious data

Measuring traveller behaviour

Smart app

pagina 20

Multimodal journey planner

Monitoring traveller behaviour

• Knowing (and affecting) traveller behaviour– Multimodal

– Door-door

– Trip purpose

• Activity-based interaction– Real time information

– Personalized information

– Gamification

• Real time questionnaires

Summary

• Big Data offers many opportunities

• Indivual mobility pattern towards aggregated mobility pat terns

to optimize design and services

• Paradigm shifts

– Passenger -> Traveller

– Trip -> Journey

– Evaluation -> Prediction