Proactive traffic control through AI and Big Data (DFF)
2022-2025: Reinforcement learning (RL) constitutes a promising solution for alleviating the problem of traffic congestion and associated greenhouse gas emissions by producing adaptive traffic signal controllers that outperform conventional systems. However, the existing solutions are merely reactive. This project aims at developing a new generation of proactive traffic control approaches by integrating into a RL framework a novel fully-Bayesian context-aware traffic prediction model that can forecast the future evolution of traffic and provide uncertainty estimates for its predictions while accounting for contextual information (e.g. about planned events, incidents and weather) traffic network flow theory and the traffic signal control actions. Read More
eMOTIONAL Cities: Mapping the cities through the senses of those who make them (Horizon 2020)
2021-2025: As the world is becoming more urbanized and cities of the future need to be people-centred, robust evidence-based knowledge on the underlying biological and psychological processes, by which Urban Planning & Design influence brain circuits and human behaviour, will be critical for policy making on urban health. Emotions are key drivers of our decisions; similarly, our choices are the conduit for our well-being and health. The eMOTIONAL Cities project aims to provide scientific evidence on how the natural and built urban environment shapes the neural system underlying human cognitive and emotional processing, with a perspective that also incorporates age, gender and vulnerable groups’ specificities.
SortedMOBILITY: Self-Organized Rail Traffic for the Evolution of Decentralized MOBILITY
2021-2024: SORTEDMOBILITY proposes a holistic approach for self-organizing management of public transport in urban and interurban areas, focusing on rail as a mobility backbone. Simulations will assess the self-organization approach in case studies in Denmark, Italy, and France. They will integrate advanced methods for passenger demand prediction and rail traffic modeling. In a close collaboration between academic and key rail stakeholders, SORTEDMOBILITY will showcase the future of railways while producing guidelines and recommendations to support future public transport systems.
BOON: Behavior Oracle for always-ON electrical mobility (DFF)
2021-2023: Electric Vehicles are increasingly adopted as part of a global transition to greener mobility, which in turn requires further charging infrastructure. This dependency is in fact circular, as new infrastructure encourages further EV adoption, and vice versa. The BOON project aims at effective planning of charging network expansion through modeling of charging demand and corresponding optimization of charging supply. For demand modeling, BOON shall develop novel data-driven Machine Learning methods for incremental learning, while accounting for inherent limitations in the observability of charging demand. On the supply side, BOON shall develop demand-responsive optimization algorithms for network expansion, involving both fixed charging stations and mobile chargers.
SHOW: SHared automation Operating models for Worldwide adoption (H2020)
2020-2024: SHOW aims to support the migration path towards affective and persuasive sustainable urban transport, through technical solutions, business models and priority scenarios for impact assessment, by deploying shared, connected, cooperative, electrified fleets of autonomous vehicles in coordinated Public Transport (PT), Demand Responsive Transport (DRT), Mobility as a Service (MaaS) and Logistics as a Service (LaaS) operational chains in real-life urban demonstrations in 5 Mega, 6 Satellite and 3 Follower Pilots taking place in 20 cities across Europe.
NOSTROMO: NEXT-GENERATION OPEN-SOURCE TOOLS FOR ATM PERFORMANCE MODELLING AND OPTIMISATION (SESAR)
2020-2022: The ATM system is composed of elements that interact with each other generating a number of properties characteristic of complex adaptive systems. NOSTROMO aims to develop new approaches to ATM performance modeling able to reconcile model transparency, computational tractability, and ease of use with the necessary sophistication required for a realistic representation of the ATM system.
Smart Mobility Management and Operation under Tradable Credit Scheme
2019-2022: Historically, inefficiencies such as congestion and vehicular emissions have been generally addressed with information provision and pricing. Recently, quantity control has been under the spotlight in transportation research, leveraging from successful applications in other economic sectors, such as the communications, energy, or environmental sectors. Limited supply is in the end a scarcity problem that can be dealt with a price instrument, a quantity instrument, or a combination of both, such as tradable credit schemes (TCS). Within a TCS system, a regulator provides an initial endowment of mobility credits to all potential travelers. In order to use a transportation system, users need to spend a certain number of permits (i.e.: tariff) that could vary with the conditions/performance of the specific mobility alternative used. The permits can be bought and sold in a market that is monitored by the regulator at a price that is determined by demand and supply interactions. In this project, we aim at leveraging the existing theoretical foundation on TCS and focus on its potential deployment and impact assessment in a realistic setting. The project will be rooted around the implementation in realistic simulation applications. Other data sources, namely from controlled experiments to be designed by the team, can be incorporated into the project to further test the developed frameworks. Read More
SHARE-MORE: SHAREd MObility REwards (EIT Urban Mobility)
2020-2020: SHARE-MORE aims to optimize the added value of car-sharing services and promote a portfolio of transport services that enable and encourage sustainable urban mobility. The effectiveness and sustainability of car-sharing integration into the bundle of transportation services will be achieved by understanding the needs of the three main stakeholders: travelers, transport authorities, and service providers, and by providing personalized incentives tailored to the needs of all three stakeholders. The incentives will be designed to increase car-sharing efficient use while contributing to the integration with the existing overall transportation system and its sustainability.
The project will balance the needed knowledge base through its consortium consisting of universities, cities, and car-sharing commercial companies to understand the underlying mechanisms of potential incentive designs, develop a specific incentive scheme, and pilot the proposed scheme within a real car-sharing service.
Integrated Learning and Optimization for Mobility and Transportation Services (ILOMYTS)
2019-2022: In this interdisciplinary project, we follow an integrated approach and build on concepts from data-driven optimization, stochastic programming, and machine learning to develop decision support with the application to transportation and mobility, in particular for bike- and car-sharing. Based on bike- and car-sharing data from the cities of Copenhagen and Munich, exploration and exploitation methods and active learning concepts will be developed to support strategic and operational decision-making with regard to capacity, inventory positioning, and resource rebalancing under uncertainty in large systems with distributed (competitive) decision making. This integrated approach requires knowledge from the disciplines of transportation and management as well as from stochastic optimization and machine learning.
Machine learning methods for transportation under uncertainty
2018-2021: The goal of the project is to develop and study effective modeling methods for Transportation under uncertainty scenarios. This is motivated by both the prevalence of uncertainty in Transportation and the widespread use of Transportation models in practice, e.g., for traffic management, planning of mobility services, and operation of Public Transport. We approach this goal through Machine Learning, namely, our proposed methods extract patterns from data and leverage them for better modeling.
Autonomous Shuttles On-Campus
2017-2020: In 2019, autonomous shuttles will start operating in DTU Lyngby campus as part of LINC: the largest test of autonomous shuttles in Denmark. The goal of our project is to dynamically predict where and when passengers would like to use the shuttles for traveling on-campus using a variety of machine learning methods.
2017-2018: PRISM is about designing, implementing and testing methodologies to better predict transport demand in a city. While plenty solutions exist today for this objective, there is general consensus that, under stress scenarios (e.g. large social events, inclement weather, demonstrations, special days), those approaches are insufficient. For example, a one-way car sharing (e.g. DriveNow) or an autonomous mobility on demand service, are highly sensitive to rebalancing operations (moving vehicles to where demand is expected).
Traffic-flow & Air Quality Experiment
2017-2018: The experimental project sought to find correlations between traffic management, traffic flow and air quality by measuring the difference between the air pollution levels when cars are waiting for the traffic light to turn green compared to when they are driving through the intersection.
The project succeeded in finding as reliable correlations for NO2 and CO as can be achieved in real-life data gathering.
2016-2018: Tripod is a system that incentivizes travelers to pursue specific routes, modes of travel, departure times, ride sharing, trip making, and driving styles in order to reduce energy use. Tripod relies on an app-based travel incentive tool designed to influence users’ travel choices by offering them real-time information and rewards.
Extraordinary Queuing detection in Denmark
2016-2017: This project helped the Danish Road Directorate in identifying extraordinary queueing situation in the main roads of Denmark using probe-vehicle data provided by INRIX. The approach consists of anomaly detection algorithms that were optimized based on data provided by the DRD. The developed methodology is currently used in production in the DRD’s traffic tower to constantly monitor the conditions of the Danish road network and detect extraordinary queueing situations.
Projects under the TINV3 – Transportens Innovationsnetværk
2014-2017: The TINV3 – Transportens Innovationsnetværk (Transport Innovation Network) helps to strengthen the Danish transport and logistics industry’s competitiveness and technological leadership. This research initiative financed a series of research projects covering topics such as:
- Prediction intervals for bus arrival times using quantile regression method and deep learning.
- Multi-modal visualization for extracting mobility insights.
- Anticipating mobility disruptions caused by special events.
Evaluation of Floating Car Data for the Danish Road Directorate
2017: In 2017, the Danish Road Directorate (DRD) considered using a Floating Car Data product, which can observe and predict traffic all over Denmark. Our task was to estimate how reliable and effective the product was. We analyzed data from the product, as gathered from vehicle fleets over 61 days in February-May 2017 in 1250 strategically important road segments.