Deep learning for enhancing mobility in smart connected cities

Deep learning for enhancing mobility in smart connected cities
Antonio D. Masegosa DeustoTech, Faculty of Engineering, University of Deusto Ikerbasque, Basque Foundation for Science
DeustoTech ( is a private non-profit institution of the Faculty of Engineering at the University of Deusto for applied research in new technologies. Since 2005 DeustoTech mission is to support the ICT activity in business and society through research, the development of technologies, innovation and knowledge transfer. We focus our activity around TRLs 2-7 and articulate it into four applied fields: Industry, Mobility, Energy and Society, having a fifth, the Chair of Applied Mathematics, as a transversal activity and support for the previous four. We are characterized for working with data of heterogeneous nature, throughout its life cycle and in compliance with ethical principles and humanists who define the University of Deusto. DeustoTech is very active in the fields of deep learning and smart cities, as it proves its wide participation in H2020 projects in this areas such as TIMON (2015-2018, GA:636220 ; ), POSTLowCIT(2016-2019, GA: 2015-ES-TM-0239-S,, LOGISTAR (2018-2021, GA: 769142, or MOMENTUM (2019-2022, GA: 815069), among others. In addition to this, the centre is highly active in international forums, as ECTRI, the European research association for sustainable and multimodal mobility and has the support of national and international external collaborators.
We are looking for post-doctoral researchers with a strong background in machine learning and/or deep learning for predictive analytics. The candidate must have a proven track record on publications in journals and conferences related to these areas. Good programming skills and experience in some of the next frameworks is required: Keras, Pytorch, Tensorflow, sci-kit learn, etc. Experience in research projects it is also desirable, particularly in H2020 or similar international programmes. Researchers with experience in the application of machine learning or deep learning techniques in any of the following fields will be especially welcomed: Smart Cities, Intelligent Transportation Systems, Travel Behaviour Analysis, Traffic Demand Analysis.
  • Information Sciences and Engineering (ENG)
A large majority of European Citizens are living in urban environments. They live their daily lives in the same space, and for their mobility share the same infrastructure. Vehicle circulation in urban environments is responsible for 40% of CO2 emissions and for 70% of other emissions. In addition, transport users suffer from congestion which increases travel times, makes travel times unpredictable, increases operating costs, wastes fuel increasing air pollution. Blocked traffic may interfere with the passage of emergency vehicles or increase the chance of collisions. Congestion costs an estimated 1% of the EU total GDP or € 100 billion per annum. Smart cities and smart mobilities are two paradigms that aim at dealing with these problems in order to make a more sustainable, efficient and inclusive urban transport. The research will focus on the design and development of new deep learning architectures for different applications in smart mobility and smart cities. These new methodologies should be capable of fusing large amounts of heterogeneous data, with spatial, temporal and contextual information, in order to make predictions of different measures at a whole city level. The MSCA researcher is expected to work with complex deep learning architectures that combine different types of neural networks(LSTMs, CNN’s, Graph CNNs, etc.) that are able to deal simultaneously with heterogeneous data types (temporal series, images, text, etc.).


J. S. Angarita-Zapata, A. D. Masegosa and I. Triguero, «A Taxonomy of Traffic Forecasting Regression Problems from a Supervised Learning Perspective,» IEEE Access, In press. JCR Journal. Ranking Q1 P. Lopez-Garcia, A. D. Masegosa, E. Osaba, E. Onieva, A. Perallos, Ensemble Classification for Imbalanced Data Based on Feature Space Partitioning and Hybrid Metaheuristics, Applied Intelligence, 2019, In press JCR Journal. Ranking Q2 P. Lopez-Garcia, E. Onieva, E. Osaba, A. D. Masegosa, A. Perallos, A Hybrid Method for Short-Term Traffic Congestion Forecasting Using Genetic Algorithms and Cross Entropy, IEEE Transactions on Intelligent Transportation Systems 17(2): 557-569, 2016 JCR Journal. Ranking Q1 P. Lopez-Garcia, E. Onieva, E. Osaba, A. D. Masegosa, A. Perallos. GACE: A meta-heuristic based in the hybridization of Genetic Algorithms and Cross-Entropy methods for continuous optimization. Expert Systems with Applications, 55, 508-519, 2016 JCR Journal. Ranking Q1 X. Zhang, E. Onieva, A. Perallos, E. Osaba, V. C.S. Lee, Hierarchical fuzzy rule-based system optimized with genetic algorithms for short term traffic congestion prediction, Transportation Research Part C: Emerging Technologies, 43 (Part 1): 127-142, 2014 JCR Journal. Ranking Q1
MOMENTUM: Modelling Emerging Transport Solutions for Urban Mobility. European Union’s Horizon 2020 Programme. Topic LC-MG-1-3-2018: Harnessing and understanding the impacts of changes in urban mobility on policy making by city-led innovation for sustainable urban mobility [Grant Agreement]. Coordinating Institution: EMT Madrid S.A. Total budget consortium: 2,927,875€. Total budget host unit: 204,000€. Duration: June 2019 – May 2022. LOGISTAR: Enhanced data management techniques for real time logistics planning and scheduling. Horizon 2020 programme – MG-5.2-2017: Innovative ICT solutions for future logistics operations [Grant Agreement 769142], Funding consortium: 4.997.548,75€. Funding host unit: 811.000€. June 2018 – May 2021. Low noise and low carbon freight delivery for Postal Operators to ensure last mile connections through optimized urban and long-distance transport (PostLowCIT), CEF Transport Programme, European Comission, 2015-ES-TM-0239-S, Funding consortium: 1.033.273€, Funding host unit: 302.925€, February 2016 – December 2019. TIMON: Enhanced real-time services for optimized multimodal mobility relying on cooperative networks and open data. Horizon 2020 – MG-3.5a-2014 – Cooperative ITS for safe, congestion-free and sustainable mobility [Grant Agreement – 636220], Funding consortium: 5.605.213 €, Funding group: 943.750€, June 2015 – November 2018.


The research posed in this thesis is located on the intersection among two knowledge areas, Intelligent Transportation Systems and Artificial Intelligence. Below, we briefly describe these areas: Intelligent Transport Systems: According to the EU Directive 2010/40/EU, Intelligent Transport Systems (ITSs) are advanced applications which without embodying intelligence as such aim to provide innovative services relating to different modes of transport and traffic management and enable various users to be better informed and make safer, more coordinated and ‘smarter’ use of transport networks. ITSs integrate telecommunications, electronics and information technologies with transport engineering. Artificial Intelligence: This discipline can be seen as a group of methodologies that aim at performing tasks that are usually assumed for humans as Natural Language Processing, Speech Recognition, Reasoning, etc. In this project, the MSCA research will mainly focus on deep learning, a subarea of Artificial Intelligence that has experienced great advances in the last decade, and it is the technology behind the main breakthroughs in Artificial Intelligence as superhuman performance in object detection in images or in speech recognition.
The two main disciplines that combine this proposal are Intelligent Transportation Systems and Artificial Intelligence. The background knowledge related to Artificial Intelligence will be provided by the scientist in charge, Antonio D. Masegosa. Dr Masegosa, awarded as IKERBASQUE Research Fellow, has a strong experience on Artificial Intelligence, particularly on metaheuristics and fuzzy systems, fields in which he has been working since he started his PhD in 2006. During this period he has published three books, 18 JCR indexed papers and more than 20 articles in both national and international conferences. The required expertise about Intelligent Transport Systems will be provided by other well-known researchers in DeustoTech. Concretely, two excellent researchers in this field as Dr Asier Perallos and Dr Enrique Onieva. Dr Perallos dean of the faculty of engineering. He has led more than a dozen of projects in this area and published nearly 20 JCR papers. In turn, Dr Onieva has also an extensive experience in this field where he accumulates more than 20 papers in JCR journals. Furthermore, the interdisciplinarity will be also strengthened through the collaboration in on-going H2020 projects as LOGISTAR or MOMENTUM.


The research proposed in this project is aligned with the H2020 societal challenge on Smart, Green and Integrated Transport, especially with two of its key objectives: – Better mobility, less congestion, more safety and security. As will be explained below, the research done in this thesis could contribute to reducing congestion and road traffic accidents. Global leadership for the European transport industry. The research proposed here would help to improve the design, planning and reliability of European distribution networks, and therefore to the global leadership of the European transport industry. – Furthermore, it is also well aligned with the societal challenge on Energy and Transport of the upcoming European Framework Programme, Horizon Europe; and with the United Nation’s Sustainable Goals 11th and 13th on Sustainable Cities and Communities, and Climate Action, respectively.


The research proposed in this offer is aligned with on-going H2020 projects in DeustoTech as LOGISTAR or MOMENTUM. In these project participate companies as Nestle UK, Pladis, Ahlers, Software AG, Aimsun or Nommon among others, and cities as Madrid, Leuven, Regensburg or Thessaloniki. The candidate will be in close cooperation with the researchers that are working on these projects so he/she will interact with these stakeholders which will provide valuable feedback to him/her, and which will follow the results of this investigation closely. This will for sure help to increase the chances of transferring the outcomes of the research accomplished in this project.


These advances with respect to the state-of-the-art foreseen in this project would have a high impact on urban mobility. Some of the main impacts are the following: – Improve design, planning and reliability of urban distribution networks. Travel times among hubs, customers, factories, providers, warehouses, etc. are pivotal in the design and planning of last mile distribution networks. The research posed here will allow us to carry out these tasks using more precise and rich information about travel times, which would result in higher reliability. – Reduce traffic congestion. Routes passing through road stretches with a higher probability of congestion are usually subject to lower reliability in travel times. If users are better informed about these facts, they will tend to avoid these routes. – Reduce traffic accidents. The stress motivated by bad travel planning or unexpected road conditions is an important cause of road accidents. The systems developed in this thesis would help the users to do better planning and to be aware of possible delays. – Reduce urban pollution. One of the aspects foreseen in this project is the use of deep learning to predict urban pollution, which will contribute to alert in advance urban authorities, helping them to take the best possible actions to mitigate high-pollution scenarios.


The main innovative aspects of the hosting offer are the following: Analysis of Open Data and Volunteered Geographic Information to take advantage of this publicly available data and provide valuable insights to citizens. Use of new Big Data technologies to deal with the vast amount of data available nowadays in cities, particularly due to the increasing digitalization of cities, thanks to the popularity of the smart city concept. Development of advance Deep Learning architectures capable of dealing simultaneously with information of a diverse nature as images, trajectory data, text, etc. Analysis of new forms of mobility as car-sharing, car-pooling, hailing services, bike-sharing services, etc. in order to help the cities understand what could be its impact on the mobility patterns of citizens.


One of the University of Deusto’s key duties is to be fully aware of problems within the institution itself and the society we live in. For this reason, it should take specific steps to boost integration and real equality of opportunity for people with specific support needs. The timely specific action is required to enable them to enter higher education in equal conditions and ensure their full integration in the university community. DeustoTech, as one of its institutions, is included in this service of social action and inclusion. The main aims consist of achieving full normalisation, equal opportunities and gradually adopting the steps needed to ensure that the University of Deusto is an inclusive educational institution. Furthermore, the University of Deusto provides them with guidance and support on the transition to the labour market jointly with special job centres and companies at large.