Institutional Repository of Vilnius University
Technologies for developing travel routing apps /
Abstract
dc:descriptionBicycle navigation applications often estimate travel time using fixed speeds, even though cycling speed varies considerably across road segments and throughout the day due to infrastructure, elevation, intersections, and rider behaviour. This thesis presents the design and implementation of a bicycle routing application that utilises historical cycling speed observations to produce routes that reflect real riding conditions. The system consists of an Android client implemented in Kotlin, backend API implemented in Go and a hosted Valhalla service. Map visualisation is provided using MapLibre, while destination search is supported via the Photon geocoder. Whenever a user starts a journey, the client records timestamped location and speed samples, which are uploaded to the backend and map matched to directed road segments using Valhalla’s trace attributes service. Valid observations are filtered, grouped into recurring time intervals, and stored in a partitioned dataset for subsequent aggregation into segment level speed profiles, including fallback strategies for sparse data. The proposed aggregation and fallback methods are evaluated using publicly available dataset of bicycle speed observations. The routing approach is validated through a simulation experiment demonstrating that speed aware routing can favour a longer path with lower expected travel time compared to baseline routing. Overall, the work shows how open map data and open source routing tools can be combined with crowd sourced speed measurements to improve bicycle route generation.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Maciūnas, Adrijanas Juozas,
Subjects
dc:subject × 1Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- eng
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD277690951&prefLang=en_US