LiRA project

The LiRA project is a proof-of-concept attempt to assess road conditions based on in-vehicle sensors on a city scale. The overall LiRA vision was to revolutionise the management of the road infrastructure by delivering a system that (i) uses car data for monitoring roads on a network level; and (ii) integrates with standard/existing measurement methods that are still employed for maintenance decisions.

The project was initiated in early 2019 (, with financial support from the Innovation Fund Denmark, with four partners:

  1. GreenMobility (GM), a car-sharing company operating a fleet of 400 Renault Zoe cars (100 % electric) within the greater Copenhagen area. Some of the GM cars were retrofitted with an IoT sensing platformthat collected readings from the native in-vehicle sensors, and also measured GPS location and 3D accelerations. These readings served as basis for the project.
  2. the Danish Road Directorate (DRD), a public authority responsible for operating, maintaining, and further developing the Danish highway network.
    DRD owns and operates several specially built vehicles for measuring road conditions. These vehicles provide a reference condition assessment to that
    inferred from the GMcars. 
  3. Sweco Denmark, a consulting company that owns, develops, and operates the current pavement management system for Copenhagen municipality. Thus, Sweco is the link for integrating the LiRA project outputs with an existing PavementManagement Systems.
  4. the Technical University of Denmark (DTU), responsible for database structuring and development ofmachine learning and physical models for data interpretation.

In general terms, LiRA focused on delivering condition information related to three different categories: (i) safety, (ii) comfort, and (iii) durability. The growing interest in emissions fromthe transportation sector has made data about energy consumption very attractive and relevant. For this reason, LiRA targeted “energy consumption” as a fourth category (iv).


Asmus Skar, Anders M. Vestergaard, Thea Brüsch, Shahrzad Pour, Ekkart Kindler, Tommy Sonne Alstrøm, Uwe Schlotz, Jakob Elsborg Larsen, Matteo Pettinari (2023). LiRA-CD: An open-source dataset for road condition modelling and research. Data in Brief, Volume 49.

Asmus Skar, Natasja Ringsing Nielsen, Matteo Pettinari and Eyal Levenberg (2023), Towards infrastructure energy labelling utilizing data from a connected fleet of electric vehicles. Transportation Research Procedia, Volume 72, Pages 2676-2683

Skar A, Vestergaard A, Pour SM, Pettinari M. (2023). Internet-of-Things (IoT) Platform for Road Energy Efficiency Monitoring. Sensors, volume 23(5):2756.

Asmus Skar and Eyal Levenberg (2023), Road Profile Inversion from In-Vehicle Accelerometers. Journal of Transportation Engineering, Part B: Pavements, Volume 150, Issue 1.

Eyal Levenberg, Asmus Skar, Shahrzad M. Pour, Ekkart Kindler, Matteo Pettinari, Milena Bajic, Tommy Alstrøm and Uwe Schlotz (2021), Live Road Condition Assessment with Internal Vehicle Sensors. Transportation Research Record, 2675(10), 1442–1452.

Zhao Du, Asmus Skar, Matteo Pettinari and Zhu, Xingyi (2023), Pavement friction evaluation based on vehicle dynamics and vision data using a multi-feature fusion network. Transportation Research Record, 2677(11), 219-236.

Eyal Levenberg (2023), Estimating the tire-pavement grip potential from vehicle vibrations. Transportation Research Record, 2677(7), 237-248.

Shahrzad M. Pour, Niels Skov Dujardin, Matteo Pettinari (2022), “Data collection”, Book chapter in “PIARC Literature review use of big data for road condition monitoring”, accepted

Milena Bajic, Shahrzad M Pour, Asmus Skar, Matteo Pettinari, Eyal Levenberg, Tommy Sonne Alstrøm, Road Roughness Estimation Using Machine Learning. (2021) arXiv preprint arXiv:2107.01199

LiRA webinar

Student projects within LiRA

Master (MSc) thesis

Markus Berthold, Live Road Assessment based on modern cars: A prototypical geographic information system for road maintenance planning, MSc Thesis 01/2020, DTU Compute,

Thomas Bech Madsen, LiRA: A Microservice Architecture for Collecting, Processing, Accessing and Visualizating Geospatial Data MSc Thesis 01/2021, DTU Compute,

Mike Glisby, LiRA PMS: A Pavement Management System Exploiting Online Road Condition Data in an Efficient Way, MSc Thesis 07/2021, DTU Compute,

Andreas Soelberg, A Method for Integrating and Organizing Software Projects Comprised of Independent Subprojects and Data Sources, MSc Thesis 02/2022, DTU Compute,

Jonas Gottlieb Svendsen, Johan Bloch Madsen, Road condition assessment using deep learning, MSc Thesis, 2021, DTU Compute.

Daniel Alejandro Campos Rivera, Map-matching GPS data to roads in Denmark by using car sensor information, MSc Thesis, 2021, DTU Compute.

Nicola Pleth, Road Damage Detection using Deep Learning with car sensor data, MSc Thesis, 2022, DTU Compute.

Jun Chen, A Scalable Architecture for Storing, Processing, Retrieving and Visualizing Road Condition Data, MSc Thesis 07/2022, DTU Compute.

Aayush Dhakal, An architectural framework for deploying machine learning to production, MSc Thesis 01/2022, DTU Compute,

Mads Finnerup Nielsen, LiRA: A Software Architecture for Deploying Predictive Machine Learning Models, MSc Thesis, In progress (expected in 02/2023), DTU Compute.

Bachelor (BSc) thesis

Jonathan Drud Bendsen, LiRA Map: A Cloud-based Geo-Information System for Road Maintenance, Bachelor thesis, BSc Project 05/2020, DTU Compute,

Alexander Faarup Christensen, A Flexible Framework for Generic Data Collection from Vehicles, BSc Thesis 06/2020, DTU Compute,

Daniel Kuzin, A Web Application for Validating Georeferenced Car Data by Domain Experts, BSc Thesis 06/2021, DTU Compute,

Martin Daniel Nielsen, Mads Dyrved Møller, A Method for Determining the Road Gradient Using Data Collected from Vehicles, BSc Thesis 08/2021, DTU Compute,

Muse Ali, Elastic Scaling of a Modularized Microservice Architecture: A Use Case on a Big Data Pipeline for Road Data Collection, BSc Project 02/2022, DTU Compute,

Nathan Maire, Machine Learning in LiRA: Aggregation and Visualization of Road Condition Data, BSc Thesis 06/2022, DTU Compute,

Sergi Doce, Understanding Road Condition and Car Data: A Flexible Framework for the Visualization of Spatio-Temporal Data, BSc Thesis 06/2022, DTU Compute,