TRAINING PROGRAMME

Advancing AI & LLM Integration for universities and academics

Empty space, drag to resize
A hands-on programme for master's and PhD students to use LLMs as a coding partner, turning raw mobility datasets into real analysis, even without prior data-science experience.
COURSE OVERVIEW

Bridging the gap between urban infrastructure and artificial intelligence.

Master's or PhD students in transport planning, data science, engineering, or urban studies already use LLM tools daily but have never applied them systematically to professional mobility data tasks. They have programming experience in at least one language and foundational knowledge of statistics, but may lack hands-on data analysis experience with Python libraries such as pandas.

The course bridges a specific gap: students who can program but have never used data analysis libraries in practice. Through LLM-assisted workflows, students learn to turn raw mobility data into insights by leveraging LLMs to generate, explain, and debug the data analysis code they have not yet learned to write on their own.

Training Format

Hybrid: online and in-person workshops.

Duration

Starting in September

Hours

4-6h

Language

All training will be in English, with Portuguese and Spanish translations available.  

Certificate

A certificate from EIT Urban Mobility. is awarded upon completion of the online course(s).

Who is this course for?

Who is AIM4Mobility addressed to?

The course is aimed at master's and/or PhD Students and academic professionals in the urban mobility ecosystem. 

Required skills:
  • Programming experience in at least one language (Python, R, MATLAB, or similar)
  • Foundational statistics knowledge from coursework
  • Familiarity with the urban mobility domain

Not required:
  • Data analysis experience with Python libraries (pandas, matplotlib) — this is what the course teaches through LLM-assisted workflows
  • Advanced prompt engineering knowledge

Tool requirements:
Google Colab (provided, no local setup). Free-tier LLMs (ChatGPT, Claude, Gemini).

Interested in learning how to use LLM and AI more effectively in urban mobility?

Enjoy the early bird discount while it lasts. 50€ 25€

Course's objectives

What will the participants get from the course

Through a structured learning journey, mobiLLITY aims to:

Select and prepare

the appropriate context for an LLM-assisted data analysis task, deciding what data to provide, in what format, and why these choices affect output quality.

Evaluate LLM-generated

analytical outputs against the source data, identifying hallucinations, fabricated data patterns, statistical errors, and misleading visualisations.

Use LLMs

to explore and clean a real mobility dataset — loading, inspecting, filtering, and transforming data through LLM-generated Python code in a Colab notebook

Design 

an end-to-end LLM-assisted analytical workflow for a mobility problem, from dataset selection and context preparation through analysis to interpretation

Generate and iterate 

on Python data analysis and visualisation code using LLMs, progressively refining outputs through context adjustments rather than starting from scratch.

Produce

a complete data analysis of a mobility dataset, demonstrating context engineering, code generation, critical evaluation, and interpretation of results.

MOBILLITY METHODOLOGY

A two-part journey: learn online, then apply it in person.

Empty space, drag to resize

MobiLLITY for Universities is delivered in two parts: a self-paced online course that every participant completes, and two optional in-person workshops that participants can add once they've finished at least one online course.

MODULES

Step 1 · For everyone

The online course

Self-paced — everyone completes this part

~6htotal time
6modules
28lectures
  • Micro-lectures, quizzes and applied tasks — no live attendance required
  • Built on real mobility data: SUMPs, GTFS data, EU directives, and citizen feedback
  • Free-tier LLMs and Google Colab — no paid subscriptions or local installs
  • Certificate on completing 5 core modules + exam; optional capstone earns Gold distinction
Step 2 · Optional
Optional

The in-person workshops

Apply what you've learned, face to face

2cities
~8heach, full day
Limitedavailability
  • Prerequisite: completion of at least one MobiLLITY online course
  • Two workshops – Lisbon, Portugal (30 September 2026) and Las Rozas, Spain (12 November 2026)
  • Full-day, in person — limited availability
  • Hands-on application of your AI and data skills to real challenges
See full workshop details

MobiLLITY does not cover travel costs.

Course structure and modules

The course is organised into six modules, progressing from context engineering foundations through applied data analysis to critical evaluation and workflow design.
Total duration: ~6 hours | Lectures: 28 | Video: ~98 minutes

MODULE 1 – Context Engineering Fundations 

This module shifts the perspective of LLM interaction from simple "prompting" to "engineering." You will explore the mechanics of the Context Window, the finite memory space of a model, and learn how to prioritise information within it.

Key Themes: Understanding token limits, choosing between JSON, CSV, or Markdown for data density, and mastering Strategic Sampling (selecting the most representative data rows to define a schema without exhausting the window).

Outcome: You will treat context preparation as a formal pre-processing step rather than an afterthought.

MODULE 2 – Advanced Data Context

Focus: Handling Relational Complexity

Building on the foundations, this module tackles the challenge of teaching an LLM to understand how disparate data files "talk" to one another.

Key Themes: Crafting Data Dictionaries that act as a map for the model, and defining relational structures for multi-table datasets. You’ll learn to build workflows that allow an LLM to navigate "linked" data without losing the thread of the relationship.

Outcome: Ability to feed an LLM complex, multi-layered data environments while maintaining structural integrity.

MODULE 3 – LLM-Assisted Data Exploration & Cleaning

Focus: From Raw Data to "Ready-to-Analyse"

This is a hands-on technical module centred on Google Colab. Instead of writing every line of boilerplate code, you will use the LLM as a sophisticated pair programmer.

Key Themes: Prompting for robust data loading scripts, using LLMs to spot outliers or missing values in mobility data, and generating automated cleaning functions.

Outcome: A significant reduction in "manual labour" time, allowing you to focus on high-level data architecture and strategy.

MODULE 4 – Iterative Code Generation & Visualisation

Focus: The Feedback Loop

A first draft of code is rarely the final one. This module focuses on the Iterative Cycle—learning how to "talk back" to the model to refine its logic or fix bugs in real-time.

Key Themes: Refining LLM-generated code through context adjustments, prompting for specific visualisation libraries (like Plotly or Matplotlib), and iteratively styling charts until they meet professional standards.

Outcome: Proficiency in steering an LLM through complex analytical tasks until the output is both accurate and visually compelling.

MODULE 5 – Critical Evaluation of LLM Outputs 

Focus: Trust but Verify

Perhaps the most critical module for professional reliability. You will develop a "skeptic’s toolkit" to identify when an LLM is being overly confident but factually wrong.

Key Themes: Output Verification (comparing LLM summaries against source truths), identifying "hallucinated" data trends, and detecting statistical errors that look plausible but are mathematically impossible.

Outcome: The ability to implement systematic evaluation workflows, ensuring that LLM-generated insights are safe for business decision-making.

MODULE 6 – Capstone (optional Gold Distiction)

Focus: Synthesis and Portfolio Building

The Capstone is where theory meets a real-world project. It is designed for those seeking a Gold Distinction by demonstrating mastery over the entire pipeline.

Key Themes: Selecting a raw mobility dataset, preparing the context, automating the cleaning/analysis, and packaging the entire process into a repeatable, end-to-end analytical workflow.

Outcome: A tangible, high-quality project that proves your ability to integrate LLMs into a professional data science lifecycle.

WORKSHOPS

Apply what you've learned online, in person.

Empty space, drag to resize

Both workshops are full-day, hands-on sessions open to any MobiLLITY audience once you've completed at least one online course. The workshops are optional, and availability is limited, so register early. If you're interested, you can choose between Lisbon or Las Rozas, whichever works better for you. 

MODULES

Workshop 1 30 September 2026

Lisbon, Portugal

Developed with ANSR, Portugal's road safety authority — turning institutional mobility and road safety data into clear visual outputs and stakeholder-ready communications.

Format
In person, full day (~8h)
Participants
Limited availability
Language
Portuguese
Data used
ANSR & public road safety data
The day in four stages
  1. 1 Data preparation — cleaning and structuring data with LLM help
  2. 2 Exploratory analysis — natural-language queries to surface patterns
  3. 3 Visualisation & communication — charts and outputs for decision-makers
  4. 4 Data privacy & responsible use — redaction and what free-tier tools can see
Workshop 2 12 November 2026

Las Rozas, Spain

Teaches a general data analysis workflow, then gives mixed teams the space to apply it to their own professional role.

Format
In person, full day (~8h)
Participants
Limited availability
Language
English
Data used
Eurostat KPIs, GTFS feeds, Barcelona IRIS
A two-part day
  1. 1 Shared foundation (~4h) — everyone works the same data analysis pipeline together
  2. 2 Role-specific application (~4h) — small groups apply it to their own domain (planning, communications, policy, operations)

Brought to you by:

FACTUAL Consulting

Leading mobility innovation consultancy.

XVAL

Experts in AI solutions for urban systems and education

EIT Urban Mobility

With the support of EIT Urban Mobility, this programme bridges the gap between cutting-edge tech and practical urban challenges.

CERTIFICATE OF COMPLETION

Once you complete the course, you'll have the opportunity to receive a certificate of completion signed by the course organizers, all of whom are recognized in the mobility industry.

Interested in learning how to use LLM and AI more effectively in urban mobility?

Enjoy the early bird discount while it lasts. 50€ 25€
This project is supported by EIT Urban Mobility, an initiative of the European Institute of Innovation and Technology (EIT), a body of the European Union. EIT Urban Mobility acts to accelerate positive change on mobility to make urban spaces more liveable. Learn more: eiturbanmobility.eu

This project is supported by EIT Urban Mobility, an initiative of the European Institute of Innovation and Technology (EIT), a body of the European Union.
EIT Urban Mobility acts to accelerate positive change on mobility to make urban spaces more liveable. Learn more: eiturbanmobility.eu