---
title: Introduction to Transportation Informatics | Lecture 1, 2026 | Masaki Ito | The University of Tokyo
tags:  #transportation informatics  
author: [Masaki Ito](https://image.docswell.com/user/niyalist)
site: [Docswell](https://www.docswell.com/)
thumbnail: https://bcdn.docswell.com/page/YE9P9N1VJ3.jpg?width=480
description: This slide is Lecture 1 of the AY2026 course Transportation Informatics, taught by Masaki Ito at the Graduate School of Information Science and Technology, The University of Tokyo. In this lecture, I introduce the scope of transportation informatics and discuss how transportation is becoming a software-defined system, along with four major trends reshaping transportation today: autonomous driving, data-driven redesign of local transportation, digital mediation of travel behavior, and AI in practice.
published: April 17, 26
canonical: https://image.docswell.com/s/niyalist/Z7NRQJ-2026-04-17-155225
---
# Page. 1

![Page Image](https://bcdn.docswell.com/page/YE9P9N1VJ3.jpg)

The University of Tokyo, Grad. School of Info. Science &
Tech. / Creative Informatics
AY2026 "Transportation Informatics" Lecture 1
April 8, 2026
Yayoi Campus, I-REF Bldg., Hilobby
Introduction to Transportation Informatics
Masaki Ito
Social ICT Research Center / Dept. of Creative Informatics (concurrent)
Graduate School of Information Science and Technology
The University of Tokyo


# Page. 2

![Page Image](https://bcdn.docswell.com/page/GE8D9K44ED.jpg)

Masaki Ito
•
•
•
Associate Professor, Social ICT Research Center, Graduate
School of Information Science and Technology, The University
of Tokyo
Research Interests
–
–
Ubiquitous computing
Transportation informatics
–
–
–
–
Born in Kakegawa, Shizuoka
2002 B.A. in Environmental Information, Keio University
2009 Ph.D. in Media and Governance, Keio University (Supervisor: Prof.
Hideyuki Tokuda)
2008-2010 Research Asst. Prof., Grad. School of Media and Governance,
Keio University
2010-2013 Assistant Professor, Graduate School of Engineering, Tottori
University
2013-2019 Assistant Professor, Institute of Industrial Science, The
University of Tokyo
2019-2021 Project Lecturer, Institute of Industrial Science, The University of
Tokyo
2021-present Current position
–
Certified Transportation Manager (Passenger)
Career
–
–
–
–
•
Qualification
2


# Page. 3

![Page Image](https://bcdn.docswell.com/page/LELMWZ6V7R.jpg)

This Course Is Conducted in English
• Per the policy of the Graduate School of Information Science and
Technology, this course is now conducted in English starting this academic
year.
• Course materials are also prepared primarily in English.
• Students may submit reports and questions in either English or Japanese.
• Transportation data used in this course are primarily from Japan and
therefore often written in Japanese. While instruction will be given in
English, some ability to read Japanese and kanji will be helpful for
understanding the data.
• Selected lectures from the 2023 edition of this course are available on
YouTube. Note that some content will be revised for the current academic
year.


# Page. 4

![Page Image](https://bcdn.docswell.com/page/4JMY9V5NJW.jpg)

How to Attend the Course
• In person: Held at Hilobby, I-REF Building, Yayoi Campus
• Online: Available via Zoom. Find the Zoom link on
UTAS/UTOL (same link used each session).
• Lecture recordings: Sessions are recorded and posted by
around the weekend — usable for review or as an
attendance alternative. Note: timely posting is not
guaranteed due to possible technical issues.


# Page. 5

![Page Image](https://bcdn.docswell.com/page/PJR9G23479.jpg)

Career and Research History
2000
2008
Student,
Tokuda Lab
Research Asst.
Prof.,
Tokuda Lab
Keio University
2010
Assistant
Professor,
Dept. of
Engineering
2013
2019
Research
Associate,
Sezaki Lab
2021
Project
Lecturer,
Oguchi Lab
Tottori University
Assoc. Professor,
Social ICT
Research Center
The University of Tokyo
Ubiquitous Computing (IoT): Pervasive Environments of Computers and Sensors
Human-Computer Interaction (HCI), UI/UX, and Social Acceptance of Computing
IT-enabled Public Transit & Community Transport
Spatial / Map Information
Environmental Sensing
Transportation Planning
Pedestrian Flow
Sensing
Traffic Engineering
ITS & Traffic Control


# Page. 6

![Page Image](https://bcdn.docswell.com/page/PEXQXD4ZJX.jpg)

Ubiquitous Computing & IoT
• A world where computers are seamlessly embedded in
physical objects and the environment. Computers form
networks and support our lives without explicit user
commands.
• My interest is a software architecture for geospatial information
and navigation, enabled by the convergence of the Cyber World
and Physical World.


# Page. 7

![Page Image](https://bcdn.docswell.com/page/3EK9WXZVED.jpg)

History of Computers:
Miniaturization and Mass Adoption
Mainframe
(1950s–)
Workstation
(1980s–)
Laptop
(1990s–)
Tablet PC
(2000s–)
Minicomputer
(1970s–)
IBM System/360
UNIX, the Internet,
and more begin
Personal Computer
(1980s–)
PDA
(1990s–)
Smartphone
(2000s–)


# Page. 8

![Page Image](https://bcdn.docswell.com/page/L73W15XQ75.jpg)

BusNet: a Regional Bus/Rail Journy Planner,
Tottori University (2010–2013)
• Over 40,000 unique users per year
• Over 300,000 searches per year
• Minister for Internal Affairs Award (Industry-Academia-Govt.
Collaboration), 2009
• Minister for Internal Affairs Award, U-Japan Grand Prize
(Regional Revitalization), 2008; and more


# Page. 9

![Page Image](https://bcdn.docswell.com/page/87DKXD2WJG.jpg)



# Page. 10

![Page Image](https://bcdn.docswell.com/page/VJPKPX1XE8.jpg)

Behavioral Analysis of BusNet Users
• Revealing travel demand for public transit through big
data analysis of web and app usage
Rank
Origin
Destination
1
Tottori
Station
AEON Tottori Kita
Aeon
Tottori Kita
Tottori Station
Tottori
Station
Prefectural Office
/Red Cross
Hospital
3
4
5
目的地設定
500
450
400
350
300
Tottori Sho
(High
school) mae
AEON Tottori Kita
Tottori
Station
Tottori Sand
Dunes
250
利用数
2
出発地設定
200
150
100
50
0
0
2
4
6
8
10
12
14
16
18
20
22
24
時間帯 h
Tottori Station Bus Stop
Demand by Route Segment
Demand Distribution by Area
Boarding/Alighting Patterns by Bus Stop


# Page. 11

![Page Image](https://bcdn.docswell.com/page/2EVV2LY3EQ.jpg)

Development of an Access Log Analysis System
• Development of a web interface for intuitive analysis
– Big-data analysis through distributed processing with Hadoop
– Selected for MIC Strategic Information and Communications R&D
Promotion Programme (SCOPE)


# Page. 12

![Page Image](https://bcdn.docswell.com/page/57GLRX9YEL.jpg)

Mobility Infrastructure
in the Age of Ubiquitous Computing


# Page. 13

![Page Image](https://bcdn.docswell.com/page/4EQYV8W6JP.jpg)

Mobility Infrastructure
in the Pre-Computer Era
• Transportation developed and grew long before
computers appeared.
• Even today, computers are often seen as just an add-on
to transportation — not as a core part of it.


# Page. 14

![Page Image](https://bcdn.docswell.com/page/KJ4WM2KM71.jpg)

Autonomous Driving as Real Infrastructure
• United States
– Waymo (Google): the only fully driverless
paid robotaxi service in the world — 100,000
rides/week as of Aug. 2024
– 1,400+ autonomous vehicles tested across
multiple states, with Arizona, Texas, and
California leading deployment
– Cruise (GM), Zoox (Amazon), and Tesla FSD
are also advancing AV technology across the
country
• China
– 32,000 km of roads open for AV testing;
Beijing demo zone expanding to 3,000 km²
– Key players: Baidu Apollo Go, Pony.ai,
WeRide — expanding globally via Uber
partnership
https://www.wired.com/story/robotaxis-cruise-waymo-san-francisco/


# Page. 15

![Page Image](https://bcdn.docswell.com/page/LE1Y8MXY7G.jpg)

Uber
•
Uber could not exist without smartphones,
GPS, and real-time data. Computers are not
an add-on — they are the foundation.
•
•
•
•
2010 Founded in San Francisco
2011 Expanded to New York and Paris
2013 Launched taxi-hailing service in Tokyo
2015 Recruited 40 researchers from Carnegie
Mellon University
2015 Ride-sharing pilot in Fukuoka halted by
Ministry of Land, Infrastructure, Transport and
Tourism
2016 Partnership with Toyota
•
•


# Page. 16

![Page Image](https://bcdn.docswell.com/page/GEWGZ3N1J2.jpg)

Uber Pool: Shared Rides Optimized by
Algorithms
• Passengers walk to a pickup
point and wait — in return, fares
drop and routes become more
efficient. This trade-off is
managed entirely by real-time
algorithms.
• Short History
– Nov 2017 Pilot launched in San
Francisco as “Uber Express Pool”
– Feb 2018 Officially launched; later
merged into Uber Pool
https://techcrunch.com/2018/02/21/uber-officially-launches-uber-express-pool-a-new-twist-on-shared-rides/


# Page. 17

![Page Image](https://bcdn.docswell.com/page/47ZL1V3XJ3.jpg)

Micro-Mobility: IT-Enabled Urban Movement
• Electric scooters and shared bikes —
collectively called “micro-mobility” — are ITnative by design.
• Every trip is tracked by GPS and managed
through smartphone apps — the entire
operation is data-driven.
• Major players (US cities, as of 2025):
– Lime (dominant), Bird (reorganized after 2023
bankruptcy), Lyft (Citi Bike / docked systems), Spin
(acquired by TIER, Germany)


# Page. 18

![Page Image](https://bcdn.docswell.com/page/YJ6WL5XPJV.jpg)

Transportation Informatics
• Transportation informatics is not simply about applying IT to
transportation. It is the discipline of understanding, analyzing,
designing, and implementing transportation systems built on
IT as their foundation.
• In this course, students learn methodologies for treating
mobility and transportation as data — analyzing, designing,
and operating systems through a data-centric approach.
• Particular emphasis is placed on hands-on skills: finding
diverse datasets, extracting and shaping the relevant
elements, and visualizing them in a variety of ways.


# Page. 19

![Page Image](https://bcdn.docswell.com/page/GJ5M1YXPJ4.jpg)

Transportation is becoming
a software-defined system.


# Page. 20

![Page Image](https://bcdn.docswell.com/page/9E291GXZ7R.jpg)

What Is Changing in Transportation Today?
• More data is being generated than ever before
• Software has moved to the core of transportation
systems
• AI is now embedded in operations and analysis
• The shift: from mode-by-mode to network-wide
optimization


# Page. 21

![Page Image](https://bcdn.docswell.com/page/D7Y4ZV8NEM.jpg)

Four Major Trends Reshaping
Transportation Today
• Trend 1: Autonomous Driving and Its Impact on
Transportation Systems
• Trend 2: Data-Driven Redesign of Local Transportation
• Trend 3: Digital Mediation of Travel Behavior
• Trend 4: AI in Practice


# Page. 22

![Page Image](https://bcdn.docswell.com/page/VENY3MQVJ8.jpg)

Trend 1: Autonomous Driving and Its
Impact on Transportation Systems


# Page. 23

![Page Image](https://bcdn.docswell.com/page/Y79P9NDVE3.jpg)

City of Tomorrow with Autonomous Vehicles
(Drive Sweden)
•
A vision of how autonomous vehicles will
transform our cities
– From streets designed for cars to spaces designed for
people
•
•
•
•
•
•
Road signs become unnecessary
More efficient road use means wider sidewalks
No need for parking lots in city centers
Vehicles waiting at the station — no waiting
for passengers
Autonomous truck platoons for efficient freight
movement
Scheduled loading and unloading reduces
parking demand
https://www.youtube.com/watch?v=WmYsWYDQxuI


# Page. 24

![Page Image](https://bcdn.docswell.com/page/G78D9K347D.jpg)



# Page. 25

![Page Image](https://bcdn.docswell.com/page/L7LMWZVVJR.jpg)

Impact 1: From Private Cars
to Shared, Integrated Mobility
• Autonomous vehicles may shift mobility from ownership to
access.
• In that case, the main unit is no longer the privately owned
car, but the mobility service.
• This connects naturally to the logic of MaaS:
– seamless connection between rail, bus, and on-demand vehicles
– unified booking, payment, and navigation
– flexible first/last-mile services
• In this view, autonomous vehicles become part of the publictransport ecosystem rather than a simple replacement for
private cars.


# Page. 26

![Page Image](https://bcdn.docswell.com/page/4EMY9VGNEW.jpg)

Impact 2: From Vehicles
to Software-Defined Mobility Platforms
• Autonomous vehicles depend on software, connectivity,
sensors, maps, and cloud-based coordination.
• This means that transportation increasingly becomes a
platform-based system.
• The core value shifts from the vehicle itself to:
– real-time data collection
– fleet coordination
– routing and dispatch
– integration with traveler information and payment systems
• In this sense, autonomous mobility is part of a broader
transition toward software-defined transportation systems.


# Page. 27

![Page Image](https://bcdn.docswell.com/page/PER9G2Z4J9.jpg)

Impact 3: From Vehicle Automation
to Urban System Reorganization
• Autonomous mobility may reshape the city itself.
• If shared autonomous services reduce the need for long-term
parking, urban land can be used differently.
• Impacts include: parking demand, curbside use, street design,
and transit hubs.
• The impact is not limited to passenger transport:
– logistics and last-mile delivery may also be integrated into the same digital
coordination logic
• The larger issue is system optimization:
– matching demand and supply more dynamically
– coordinating passenger and freight movement together


# Page. 28

![Page Image](https://bcdn.docswell.com/page/P7XQXD9ZEX.jpg)

TESLA
•
An electric vehicle company founded by Elon
Musk
– Founded in 2003
•
Autonomous driving hardware comes standard
•
Algorithm improves through real-world driving
data from users
Customization and ordering available online
•
– Cameras, ultrasonic sensors, and radar for environment
recognition
– Autopilot feature provided
– Not fully autonomous yet, but full self-driving capability
planned
– New features added via over-the-air software updates


# Page. 29

![Page Image](https://bcdn.docswell.com/page/37K9WXPV7D.jpg)

Tesla FSD (Full Self-Driving)
• Starting with FSD v12.5.5, released in 2024, Tesla
switched to an end-to-end AI approach, significantly
improving autonomous driving performance.
– Previous approach: camera and sensor inputs were interpreted by rulebased programs to determine driving actions
• e.g., "The arrow signal ahead points right, no obstacle in front — enter the
intersection and turn right at current speed"
– E2E approach: a neural network directly learns the mapping from sensor
inputs and spatial context to driving actions, without intermediate rulebased interpretation. Training data is collected from a large fleet of Tesla
vehicles on the road.


# Page. 30

![Page Image](https://bcdn.docswell.com/page/LJ3W154QJ5.jpg)

Trend 2: Data-Driven Redesign of
Local Transportation


# Page. 31

![Page Image](https://bcdn.docswell.com/page/8JDKXDQWEG.jpg)

Data-Driven Redesign of Local
Transportation
• Local transportation is no longer only a matter of maintaining
existing bus and rail services.
• It is increasingly a matter of redesigning mobility systems under
demographic, financial, and operational constraints.
• In many regions, the challenge is not how to optimize one operator,
but how to reorganize the entire local mobility ecosystem.
• This includes:
– bus services
– rail connections
– taxis and on-demand services
– first/last-mile access
– coordination among public and private actors


# Page. 32

![Page Image](https://bcdn.docswell.com/page/VEPKPXLX78.jpg)

“Transportation Desert” Elimination and FullScale Re-Design Project (Ministry of Land,
Infrastructure, Transport and Tourism・MLIT)
https://kotsu-kuhaku-r8.jp


# Page. 33

![Page Image](https://bcdn.docswell.com/page/27VV2LQ37Q.jpg)

Re-Design: Four Key Targets
• ① Eliminating “Transportation Deserts”
– Supporting locally-led systems such as Japanese-style ride-sharing and demandresponsive transit
• ② Promoting Consolidation and Collaboration
– Supporting joint operations among transport operators (buses, taxis) to improve
efficiency
• ③ Advancing Digital Transformation (DX) in Local
Transportation
– Supporting data and system integration for advanced services through digital
technology
• ④ Developing Mobility Experts and Organizational Capacity
– Supporting local governments in data-driven reviews of regional transportation with
stakeholders


# Page. 34

![Page Image](https://bcdn.docswell.com/page/5JGLRXWY7L.jpg)

Local Transportation Is a Governance
Problem as Much as a Technical One
• Local transportation is rarely managed by a single actor.
• In practice, it involves:
– local governments
– transport operators (railway, bus, taxi, etc.)
– police
– residents and community representatives
• Redesign is not only a technical problem but also a
problem of coordination, negotiation, and public
decision-making.


# Page. 35

![Page Image](https://bcdn.docswell.com/page/47QYV836EP.jpg)

From Experience-Based Operations to
Data-Driven Redesign
• Local transportation has often relied on experience,
intuition, and practical judgment.
• This knowledge remains valuable and often indispensable.
• But redesign now requires decisions that are more
transparent, shareable, and responsive to change.
• Data does not replace experience.
• It makes judgment more visible, testable, and
discussable.


# Page. 36

![Page Image](https://bcdn.docswell.com/page/KE4WM21MJ1.jpg)

Example of a Recent Policy Initiative in Japan
• The Japanese
government is moving to
make local transportation
data easier to share
among operators and
local governments, in
order to support route
redesign and the
sustainable provision of
local mobility services.
https://www.nikkei.com/article/DGXZQOUA206FO0Q6A220C2000000/


# Page. 37

![Page Image](https://bcdn.docswell.com/page/L71Y8MGYJG.jpg)

Why Data Matters in Local
Transportation Redesign
• Data helps reveal problems of local transportation:
– Coverage gaps — areas and users left without adequate service
– Service gaps — mismatches in timing, frequency, and connections
– Demand gaps — where supply no longer matches actual travel needs
• Data also creates a shared basis for discussion.
– When stakeholders work from the same maps and data, discussion can
move from opinion to problem-solving.
• The goal is to support informed decision-making, not
only analysis.


# Page. 38

![Page Image](https://bcdn.docswell.com/page/G7WGZ3K1E2.jpg)

Data as a Starting Point for Dialogue —
Kumamoto
• Representatives from
Kumamoto City, bus
operators, and
university researchers
meet regularly to plan
and discuss field trials
in the Kumamoto area.
• Shared data on the
screen gives everyone
a common ground to
work from.
41


# Page. 39

![Page Image](https://bcdn.docswell.com/page/4JZL1VZXE3.jpg)

17年7月
17年9月
17年11月
18年1月
18年3月
18年5月
18年7月
18年9月
18年11月
19年1月
19年3月
19年5月
19年7月
19年9月
19年11月
20年1月
20年3月
20年5月
20年7月
20年9月
20年11月
21年1月
21年3月
21年5月
21年7月
21年9月
21年11月
22年1月
22年3月
22年5月
22年7月
22年9月
22年11月
23年1月
23年3月
23年5月
23年7月
23年9月
23年11月
24年1月
24年3月
24年5月
24年7月
24年9月
24年11月
25年1月
25年3月
25年5月
25年7月
25年9月
25年11月
Growth in Public Transit Open Data
Providers in Japan (2017–)
オープンデータ提供事業者数
800
700
600
500
400
300
200
100
0


# Page. 40

![Page Image](https://bcdn.docswell.com/page/YE6WL5ZPEV.jpg)

The GTFS Format
• A globally adopted open data standard for public transit
• A file format that bundles all information needed for transit trip
planning: stops/stations, routes, schedules, and fares
Stops/Stations + Routes
Schedules
Fares


# Page. 41

![Page Image](https://bcdn.docswell.com/page/GE5M1YWPE4.jpg)



# Page. 42

![Page Image](https://bcdn.docswell.com/page/97291GQZJR.jpg)

Trend 3: Digital Mediation of Travel
Behavior


# Page. 43

![Page Image](https://bcdn.docswell.com/page/DJY4ZVWN7M.jpg)

Transit Apps in Japan
駅すぱあと
駅探 乗換案内
ジョルダン 乗換案内
Yahoo!乗換案内
NAVITIME
Google Maps
Apple Maps


# Page. 44

![Page Image](https://bcdn.docswell.com/page/V7NY3M9VE8.jpg)

Digital Mediation of Travel Behavior
• Travel behavior is increasingly shaped through digital
interfaces.
• People are no longer interacting only with vehicles, roads
and physical signboards.
• They are also interacting with software, rankings,
recommendations, alerts, and platform logic.


# Page. 45

![Page Image](https://bcdn.docswell.com/page/YJ9P9N2V73.jpg)

The Growing Role of Smartphones
in Travel Behavior
• Why Did It Fail? Transit Apps on a
Snowy Day
– Misled by the app — missed the bus three
times
– The bus I planned to take disappeared
from the app
– Kept searching on the taxi app — no luck
at all
• → Users now expect apps to work
not just in normal conditions, but
during emergencies too
– “Bad weather means disruption” is no
longer an acceptable excuse
NHK NEWS Web, January 19, 2016
48


# Page. 46

![Page Image](https://bcdn.docswell.com/page/GJ8D9K54JD.jpg)

SEO Thinking in Public Transportation
Nikkei MJ, October 19, 2015 / Interview with
the President of Keihan Electric Railway
When planning a train journey, most
passengers turn to a transit app on their
smartphone.
No matter how much a railway promotes its
appeal, most riders simply take the first
option the app shows. Ranking at the top of
transit search results is becoming the single
most important factor in winning passengers.
This is something we cannot ignore. That is
why we work hard to shave off even one or
two minutes of travel time.
49


# Page. 47

![Page Image](https://bcdn.docswell.com/page/LJLMWZYVER.jpg)

Digital Information Does Not Only
Support Travel — It Shapes It
• Digital tools do more than provide information.
• They actively influence:
– which mode people choose
– which route they take
– when — or whether — they travel
• Mobility platforms don’t just respond to demand. They
can reshape it.


# Page. 48

![Page Image](https://bcdn.docswell.com/page/47MY9VNN7W.jpg)

From Guidance to Steering
• Digital systems can now actively steer travel behavior at
scale.
• New possibilities:
– peak spreading, congestion mitigation, mode shift, demand management
• New questions:
– Who decides what gets recommended — and on what basis?


# Page. 49

![Page Image](https://bcdn.docswell.com/page/P7R9G2D4E9.jpg)

石村怜美, 梶原康至, 太田恒平: 「乗換検索サービス
の経路選択データを用いた公共交通の経路選択行
動分析」, 第49回土木計画学研究発表会, 2014.
• x


# Page. 50

![Page Image](https://bcdn.docswell.com/page/PJXQXD1Z7X.jpg)

• x


# Page. 51

![Page Image](https://bcdn.docswell.com/page/3JK9WX2VJD.jpg)

太田恒平, 渡部啓太, 小竹輝幸, 梶原康至: 「カーナビ
が 経路選択を左右する」, 第53回土木計画学研究発表
会, 2016年.
• x


# Page. 52

![Page Image](https://bcdn.docswell.com/page/LE3W15YQE5.jpg)

Why This Matters for Transportation
Informatics
• Transportation design must now include:
– information design, interface design, ranking logic, incentive design,
platform governance
• Key insight:
• Transportation informatics is also about how digital
systems shape perception, choice, and movement.


# Page. 53

![Page Image](https://bcdn.docswell.com/page/8EDKXD5W7G.jpg)

Trend 4: AI in Practice


# Page. 54

![Page Image](https://bcdn.docswell.com/page/V7PKPXGXJ8.jpg)

AI in Practice
•
•
•
•
AI in transportation is no longer only a research topic.
It is increasingly used in practical, task-specific ways.
In most cases, AI does not replace the whole system.
It supports particular functions such as:
– prediction
– monitoring
– anomaly detection
– decision support
– communication and user assistance
• Key point:
• The most realistic role of AI today is not full automation,
• but practical support for specific transportation tasks.


# Page. 55

![Page Image](https://bcdn.docswell.com/page/2JVV2L63JQ.jpg)

OECD Report: AI in Mobility
• The mobility sector is foundational to
the EU economy — but faces growing
pressure to adapt.
• AI is emerging as a key enabler of
smarter, more sustainable, and
resilient mobility
• Three application fields examined in
depth:
– Automated driving
– Public transport
– Fleet management (freight)
• Based on literature review +
interviews with EU businesses (Dec
2024 – Apr 2025)
https://www.oecd.org/en/publications/progress-inimplementing-the-european-union-coordinated-plan-onartificial-intelligence-volume-2_3ac96d41-en/full-report/ai-inmobility_3606a201.html
February 18, 2026


# Page. 56

![Page Image](https://bcdn.docswell.com/page/5EGLRXNYJL.jpg)

Application Areas of Generative AI in
Transportation Planning
•
Descriptive Tasks for Data Fusion and Analytics
•
Predictive Tasks in Transportation Planning
•
Generative Tasks: Data Synthesis and Scenario Generation
•
Simulation Tasks for Mixed Traffic Environments
– Collecting, processing, integrating, and analyzing transportation data to extract actionable insights
– Understanding system status, identifying patterns, and detecting anomalies
– Forecasting traffic flow, arrival times, travel demand, and infrastructure performance using historical and
real-time data
– Potential to capture dynamic, multi-dimensional patterns beyond regression models and rule-based
simulations
– Creating synthetic datasets for data-scarce scenarios; running hypothetical simulations
– Addressing: data collection costs, privacy concerns, and rare events
– Enabling complex, high-fidelity traffic simulations
•
Including mixed scenarios with human-driven vehicles (HVs) and autonomous vehicles (AVs)
•
Trustworthiness in GenAI-Based Transportation Systems
•
Source: Da, L., Chen, T., Li, Z., Bachiraju, S., Yao, H., Li, L., Dong, Y., Hu, X., Tu, Z., Wang, D., et al.: Generative AI in Transportation
Planning: A Survey, arXiv preprint arXiv:2503.07158, 2025.
– Six key challenges: privacy, security, fairness, accountability, explainability, and reliability


# Page. 57

![Page Image](https://bcdn.docswell.com/page/4JQYV8Q67P.jpg)

Where AI Is Already Becoming Useful
• AI is being applied in areas where transportation systems
generate large amounts of operational data.
• Realistic application areas include:
– predictive maintenance of vehicles and infrastructure
– video-based safety monitoring
– traffic and demand prediction
– dynamic routing and dispatch
– customer support and passenger information
• These applications are practical because they:
– use existing data streams
– solve narrow but important problems
– can be introduced without redesigning the whole system


# Page. 58

![Page Image](https://bcdn.docswell.com/page/K74WM2NME1.jpg)

How Well Can LLMs Answer Questions in
Transportation Planning?
• GPT-4 and Phi-3-mini evaluated
across three areas:
1. Geospatial processing capabilities
2. Domain-specific knowledge in
transportation
3. Real-world problem-solving in
congestion pricing scenarios
•
GPT-4 outperformed Phi-3-mini
across all levels: 86% on GIS
tasks, 81% on MATSim
comprehension, and 91% on
real-world transportation
decision support.
“Beyond Words: Evaluating Large Language Models in
Transportation Planning”, Ying, Shaowei, Zhenlong Li, and
Manzhu Yu, Geo-Spatial Information Science 1, 23 (2025).


# Page. 59

![Page Image](https://bcdn.docswell.com/page/LJ1Y8M5YEG.jpg)

Feeding Transportation
Data
to AI Agents


# Page. 60

![Page Image](https://bcdn.docswell.com/page/GJWGZ35172.jpg)

Person Trip Survey Data as an MCP
Server
• Downloaded the 6th
Tokyo Metropolitan
Person Trip Survey
data (Heisei 30 /
2018)
– Cross-tabulated data
from multiple
perspectives is publicly
available
• Distributed Excel files
imported into
PostgreSQL, then
exposed as an MCP
server
https://www.tokyo-pt.jp/data


# Page. 61

![Page Image](https://bcdn.docswell.com/page/4EZL1VNX73.jpg)

Transportation
Data Analysis:
A Demo
• Starting from
understanding the
entire database, then
locating and
interpreting the
needed data
– Runs autonomously for
several minutes to over
10 minutes per question
• Uses SQL, Python, etc.
to retrieve data and
generate charts
• Presents findings not
only as charts but also
as written insights
10x speed playback


# Page. 62

![Page Image](https://bcdn.docswell.com/page/Y76WL58P7V.jpg)

QGIS Control: A Demo
•
•
•
Instructed
to draw a
Shinkansen
route map
Started with
major
stations;
added all
stations
upon further
instruction
For some
reason, the
data got
corrupted
midway
25x speed playback


# Page. 63

![Page Image](https://bcdn.docswell.com/page/G75M1Y9P74.jpg)

What This Means for Transportation
Informatics
• AI does not eliminate the need to understand data, systems, and
context.
• In fact, AI becomes useful only when:
– data are well structured
– operational goals are clear
– outputs can be verified
– human judgment remains involved
• For this course, AI should be treated as a practical tool for:
– coding support
– exploratory analysis
– data cleaning assistance
– visualization support
– explanation and communication


# Page. 64

![Page Image](https://bcdn.docswell.com/page/9J291G2ZER.jpg)

Section Summary: The SoftwareDefined Transportation System
• Software-defined Transportation System
•
•
•
•
More data is being generated than ever before
Software has moved to the core of transportation systems
AI is now embedded in operations and analysis
The shift: from mode-by-mode to network-wide optimization
• Major Trends
– Trend 1: Autonomous Driving and Its Impact on Transportation Systems
– Trend 2: Data-Driven Redesign of Local Transportation
– Trend 3: Digital Mediation of Travel Behavior
– Trend 4: AI in Practice


# Page. 65

![Page Image](https://bcdn.docswell.com/page/DEY4ZVYNJM.jpg)

Transportation
Informatics


# Page. 66

![Page Image](https://bcdn.docswell.com/page/VJNY3MPV78.jpg)

Objectives and Scope of
Transportation Informatics
•
•
•
This course covers the transportation domain, which is rapidly evolving through
integration with information technology. Students will acquire fundamental and
practical skills in transportation data analysis, geospatial information processing,
traffic simulation, and transportation service design.
In transportation engineering and planning, leveraging diverse data on road and
public transit systems to achieve safer, smoother, more convenient, and efficient
transportation infrastructure and services is increasingly important. Advances in
GIS, databases, simulation, machine learning, and AI have greatly expanded and
enhanced the methods available for transportation analysis and design.
Through this course, students will explore the latest case studies and research on
transportation data collection, visualization, analysis, and societal applications.
Working hands-on with real transportation data, students will develop practical
skills in programming, data analysis tools, and AI — building a foundation of
transferable competencies applicable to both research and professional practice.


# Page. 67

![Page Image](https://bcdn.docswell.com/page/YE9P9N3VJ3.jpg)

Course Schedule (13 Lectures)
1. Introduction to Transportation Informatics
2. GIS and Spatiotemporal Databases 1
3. GIS and Spatiotemporal Databases 2
4. Public Transit Data Analysis with PostgreSQL + PostGIS + QGIS 1
5. Public Transit Data Analysis with PostgreSQL + PostGIS + QGIS 2
6. Public Transit Data Analysis with PostgreSQL + PostGIS + QGIS 3
7. AI and Data Analysis
8. Network Search and Road Traffic
9. Introduction to Microscopic Traffic Simulation: SUMO
10. Advanced Microscopic Traffic Simulation: SUMO Applications
11. Urban Transportation Planning and Data
12. AI and Transportation Simulation
13. The Future of Transportation Informatics (Discussion)


# Page. 68

![Page Image](https://bcdn.docswell.com/page/GE8D9KM4ED.jpg)

Geographic Information Systems (GIS)
• Maps are essential for understanding and
analyzing transportation.
• We will develop skills for working with
spatial data:
– Visualization
• How to represent the information you want to communicate
– Spatial operations
• e.g., Can you calculate the distance between 2 points from their
lat/lon coordinates?
• QGIS Exercise
– An open-source GIS tool
• Where to find spatial data
– Census data, mesh data, and more


# Page. 69

![Page Image](https://bcdn.docswell.com/page/LELMWZ3V7R.jpg)

Public Transit Data Analysis with
PostgreSQL + PostGIS + QGIS
• Learning SQL for spatial data management and analysis
– Managing large datasets with files or Excel alone is inefficient
– SQL is a transferable skill — applicable to personal PCs, enterprise
databases, and big data platforms
• SQL: A programming language for relational databases
– Used with PostgreSQL, Oracle Database, MySQL, Google Cloud BigQuery,
and more
• PostgreSQL + PostGIS
– Open-source RDBMS with spatial data extension
– Enables geospatial queries and analysis on transportation data


# Page. 70

![Page Image](https://bcdn.docswell.com/page/4JMY9VDNJW.jpg)

Metropolitan Transportation Census
• Conducted every five years by MLIT in the three major
metropolitan areas (Tokyo, Nagoya, Osaka) to survey the
actual usage of mass public transit (railways, buses, etc.)
• The most recent survey was conducted in FY2021.
• Survey Methods
– Up to the 12th survey: paper questionnaires distributed at stations,
returned by mail and statistically expanded (sample survey: 320,000
responses)
– 13th survey: aggregated from railway IC card data — contactless, full
census (19.15 million records)
https://www.mlit.go.jp/sogoseisaku/transport/sosei_transport_tk_000007.html


# Page. 71

![Page Image](https://bcdn.docswell.com/page/PJR9G24479.jpg)

Microscopic Traffic Simulation: SUMO
• Models individual vehicles and
places them in a simulated
road environment
– Vehicle behaviors modeled: carfollowing, lane-changing, etc.
• Reproduces traffic conditions
under specific scenarios
– Inputs: road network, traffic volume,
signal timing, etc.
• Types of traffic flow
simulators
– Also includes macro- and mesosimulators


# Page. 72

![Page Image](https://bcdn.docswell.com/page/PEXQXD2ZJX.jpg)

Generative AI and AI Agents
• We plan to use these in this course as well.
• Details will be worked out as the course progresses…


# Page. 73

![Page Image](https://bcdn.docswell.com/page/3EK9WXMVED.jpg)

Course Materials and Lecture Videos
• Slides will be shared on UTOL before each lecture.
• Lecture videos will be shared:
– After minor editing, uploaded to YouTube by around Friday
– Please use them for review before assignments


# Page. 74

![Page Image](https://bcdn.docswell.com/page/L73W152Q75.jpg)

Part of the Course Is
Open to the Public
• The parts where the
instructor is speaking
(excluding student
discussions) are made
publicly available.
– Course materials and exercise
data are also publicly available.
https://itolab.t.u-tokyo.ac.jp/education/


# Page. 75

![Page Image](https://bcdn.docswell.com/page/87DKXDZWJG.jpg)

Assignments and Grading
• Attendance comment (each lecture)
– Submit via UTOL (including today)
– Feedback and discussion during class
– Deadline: 24:00 one week later
• Midterm report
• Final report
• Grading
– Attendance 2 : Midterm 3 : Final 4
– Submitting either report prevents a grade of "not attempted"


# Page. 76

![Page Image](https://bcdn.docswell.com/page/VJPKPXDXE8.jpg)

AY2025 Midterm Report Assignment
• Prompt: Using the Metropolitan Transportation Census,
identify a transportation phenomenon you find interesting and
explain it using maps and charts.
• Length: 1,000+ characters (Japanese) or 500+ words
(English) + 2 or more figures/tables
• SQL: If you use SQL, include it in the report (appendix is fine).
Use of SQL earns bonus points.
• Generative AI: If used, describe how in the report (appendix
is fine). No penalty for any use.
• Any questions about the assignment? Please leave them as
today’s attendance comment.


# Page. 77

![Page Image](https://bcdn.docswell.com/page/2EVV2L13EQ.jpg)

AY2025 Final Report Assignment
• Length:
– 1,500+ characters (Japanese) or 750+ words (English) + 3 or more figures/tables
– Attach SQL, Python, or other code used as an appendix
• Prompt: Based on the content of Transportation Informatics
(Advanced), identify a topic that could contribute to the future of
transportation and discuss it.
• Example topics (you are free to choose your own; originality of
the topic itself is not required):
– Policy recommendations based on transportation data
– Explaining transportation phenomena using data and simulation
– Cross-regional comparison of transportation using GIS
– Survey of tools for transportation data analysis and their applicability
• Bonus points (hands-on use of IT and data is valued):
– SQL, transportation big data, GIS, traffic simulation
– Analysis combining multiple datasets
– Use of data or IT tools not covered in class
– Programming


# Page. 78

![Page Image](https://bcdn.docswell.com/page/57GLRX2YEL.jpg)

Today's Assignment
• Please share your impressions of the course and what
you hope to learn through it.
• Submit via UTOL
– Deadline: April 15 (Wed) 24:00


# Page. 79

![Page Image](https://bcdn.docswell.com/page/4EQYV8P6JP.jpg)

Preparation for Next Class
• Install QGIS
– Please install QGIS on
your own PC before the
next class.
• Version
– 4.0.0: Latest release
– 3.44.8 Long Term
Release (LTR) —
recommended for
stability


# Page. 80

![Page Image](https://bcdn.docswell.com/page/KJ4WM2YM71.jpg)

Q&A