This thesis investigates how bikeability-related attributes influence route choice in urban cycling, aiming to integrate behavioral considerations into operational routing procedures. Using GPS trajectories and questionnaire data collected from university students in Milan, a scenario-based routing framework is developed to translate link-level characteristics into perceived generalized costs. Rather than estimating a full discrete choice model, the methodology defines stylized behavioral scenarios representing different cyclist archetypes. Distance is treated as a baseline cost, while cycling infrastructure, traffic exposure, greenery, and infrastructure typology modify perceived cost under a non-negativity constraint compatible with Dijkstra’s algorithm. Observed routes are compared with scenario-generated alternatives through route-level indicators and a geometric overlap measure. Results reveal a scale-dependent pattern. Short trips show limited divergence across scenarios, while medium and long trips exhibit corridor-level substitution as preferences reshape route alignment. Infrastructure-oriented scenarios uncover extended protected spines capable of substantially increasing facility exposure with moderate detours. Traffic-averse specifications consistently reduce motorized exposure, sometimes independently of infrastructure provision. Greenery exposure remains generally low and unevenly distributed, limiting its capacity to restructure routes and occasionally producing overlaps between scenarios such as Balanced and Traffic Relaxed. Overall, the framework highlights structural trade-offs between directness, traffic avoidance, and infrastructure provision, offering a flexible tool for scenario testing in cycling planning.
Questo lavoro analizza come gli attributi legati alla bikeability influenzino la scelta del percorso nel ciclismo urbano, con l’obiettivo di integrare considerazioni comportamentali nelle procedure di instradamento. Utilizzando traiettorie GPS e dati da questionario raccolti tra studenti universitari a Milano, viene sviluppato un framework di routing basato su scenari che traduce le caratteristiche degli archi in costi generalizzati percepiti. Invece di stimare un modello completo di scelta discreta, la metodologia definisce scenari comportamentali stilizzati che rappresentano diversi archetipi di ciclista. La distanza costituisce il costo di base, mentre infrastruttura ciclabile, esposizione al traffico, verde urbano e tipologia infrastrutturale modificano il costo percepito, nel rispetto di un vincolo di non negatività coerente con l’algoritmo di Dijkstra. I percorsi osservati sono confrontati con quelli generati nei diversi scenari tramite indicatori aggregati e una misura geometrica di sovrapposizione. I risultati mostrano un pattern dipendente dalla scala: differenze limitate nei percorsi brevi, sostituzioni di corridoio nei percorsi medi e lunghi. Gli scenari orientati all’infrastruttura evidenziano dorsali protette capaci di aumentare significativamente l’esposizione a infrastrutture con deviazioni moderate. L’esposizione al verde risulta generalmente contenuta e disomogenea, con parziali sovrapposizioni tra scenari come Balanced e Traffic Relaxed. Il framework evidenza i compromessi tra direttezza, riduzione del traffico e dotazione infrastrutturale, offrendo uno strumento flessibile per l’analisi di scenari nella pianificazione ciclabile.
Understanding bikeability and route choice in urban cycling
LEON GIRALDO, JUAN FELIPE
2025/2026
Abstract
This thesis investigates how bikeability-related attributes influence route choice in urban cycling, aiming to integrate behavioral considerations into operational routing procedures. Using GPS trajectories and questionnaire data collected from university students in Milan, a scenario-based routing framework is developed to translate link-level characteristics into perceived generalized costs. Rather than estimating a full discrete choice model, the methodology defines stylized behavioral scenarios representing different cyclist archetypes. Distance is treated as a baseline cost, while cycling infrastructure, traffic exposure, greenery, and infrastructure typology modify perceived cost under a non-negativity constraint compatible with Dijkstra’s algorithm. Observed routes are compared with scenario-generated alternatives through route-level indicators and a geometric overlap measure. Results reveal a scale-dependent pattern. Short trips show limited divergence across scenarios, while medium and long trips exhibit corridor-level substitution as preferences reshape route alignment. Infrastructure-oriented scenarios uncover extended protected spines capable of substantially increasing facility exposure with moderate detours. Traffic-averse specifications consistently reduce motorized exposure, sometimes independently of infrastructure provision. Greenery exposure remains generally low and unevenly distributed, limiting its capacity to restructure routes and occasionally producing overlaps between scenarios such as Balanced and Traffic Relaxed. Overall, the framework highlights structural trade-offs between directness, traffic avoidance, and infrastructure provision, offering a flexible tool for scenario testing in cycling planning.| File | Dimensione | Formato | |
|---|---|---|---|
|
2026_03_Leon_v1.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: testo tesi
Dimensione
2.17 MB
Formato
Adobe PDF
|
2.17 MB | Adobe PDF | Visualizza/Apri |
|
2026_03_Leon.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: testo tesi versione finale
Dimensione
4.06 MB
Formato
Adobe PDF
|
4.06 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/252523