Description du sujet de thèse

Motivation of PhD
Our society is more and more conscious of the contribution of current mobility modes to the climate crisis. This is why innovative low-carbon mobility solutions are being promoted by decision-makers and increasingly adopted by citizens. It is, for instance, expected that Electric Vehicles (EVs) will account for 70% of sold vehicles by 2030. The EU Commission, with its Fit 55 plan, even envisions a ban on the sale of new petrol and diesel cars as early as 2035. Meanwhile, adoption of micromobility modes is increasing significantly. Micromobility is an umbrella term used to describe the category of transportation using non-conventional battery-powered vehicles aimed at shrinking the physical and environmental footprint required for quickly moving people over relatively short distances. With micromobility, urban transportation modes have diversified very quickly. The challenge for cities encompasses organization and planning of public space and promotion of active mobility for health purpose given the passivity of some micromobility modes (e-scooters in particular). The co-existence of these modes in shared spaces cause various kinds of inconvenience for other users (people in wheelchairs, walking with a baby in a stroller, or elderly people) and alters the perception of safety which can lead vulnerable people to be more sedentary. Beyond the perception, it is attested that the number of accidents due to e-scooters is constantly increasing. It is therefore crucial to monitor the use of these micromobility modes by collecting information in a dynamic and non-intrusive way and then make recommendations for safer shared spaces and physical activity.
Proposed work during PhD
Three main tasks are envisioned for this thesis:
a)- City-wide mobility model: This task aims at developing a dynamic network model for multimodal mobility over a city. For this purpose, our starting point will be the recent works by the team which developed a large-scale mobility model to characterize the daily movement of people in an urban network. This model is based on the modeling of people's mobility between their place of residence and 5 categories of destinations (work, schools, etc.). It generates a graph with nodes (origins and destinations) and also their interconnections through the origin-destination matrix that characterizes: directions, weights and temporal profile of the connections between nodes. The model simulates the movement of people at an aggregate level (no distinction of individuals, no information on the routes connecting origin and destination), Pratap et al. 2022. It has been used to control epidemics propagation while preserving the territory productivity, Niazi et al. 2021.
For monitoring multi-modal mobility, we will divide the city in cells. Each cell will define a node of the mobility networks. Each node will have several states representing the number of users for each mobility mode. Transition can be done from one mode to another. Therefore, there will be a dynamic for mobility mode in each node. Each node will interact with its neighbors. Two nodes will be adjacent if there is at least one mode from which people can jump from one node to another. The graph is expected to be large and dense with weights related to mobility between nodes. The originality of this task rests in the finer grain of the proposed description and the accurate distinction between the possible transportation modes, including cars, public transportation and micromobility.
b)- From discrete to continuous: here we will develop a dynamic continuous counterpart to the discrete city-wide network of the previous task by using graphons (Ruiz et al, 2021) and/or continuation (Nikitin et al. 2021). The city-wide network from the previous task is equipped with a dynamics for the evolution of the shares of the mobility modes. We expect that this dynamics will feature diffusion and transport terms: therefore, the dynamics belong to a class that we are able to treat by continuation and graphon methods (or a combination of both). While the apparent geographical interpretation is conductive to continuation methods, public transportation (such as tramways in the Grenoble case study) effectively introduces long-range connections and counters the inherent sparsity of the micromobility diffusion.
c)- Network estimation: The aim is to develop techniques for state estimation for providing a cartography of transportation modes usage. The estimation problem here consists in: (i) the estimation of connection weights (probabilities) in the context of dynamic network models; (ii) estimate the states of the nodes or an aggregated states in order to provide a cartography of co-existence of transportation modes. Connection weights will be treated as functional data. The question here is therefore to estimate the tensor of weights (or connection probabilities; tensor here means multiway array from partial measurements provided by people using mobility modes detectors (Taia-Alaoui et al. 2022) or other sensors like loop detectors for vehicles and bicycles (the latter are available in the city of Grenoble and our team has acquired experience with dealing with similar data in building the GTL and GTL-Villle platforms). This task will include methods for dynamic network completion using graphon approximation, which will be acutely needed to cope with partial observations. These methods will be based on related methods proposed for collaborative filtering, such as (Shah and Lee, 2018). Furthermore, we will develop techniques to detect (despite noisy and possibly patchy data) potential conflict zones, where conflict and safety issues originate from high rates of penetration of micromobility modes such as e-scooter (Dozza et al. 2022).
References
Dozza, M., Violin, A., and Rasch, A. A data-driven framework for the safe integration of micro-mobility into the transport system: Comparing bicycles and e-scooters in field trials, Journal of Safety Research, Vol. 81, 2022, Pages 67-77.
Niazi, M.U.B., Canudas de Wit, C., Kibangou, A.Y., and Bliman, P.-A. (2021). “Optimal control of urban human mobility for epidemic mitigation”, 60th IEEE Conference on Decision and Control (CDC), Austin, TX USA.
Niazi, M.U.B, C. Canudas-de-Wit, C., and A. Kibangou. Average state estimation in large-scale clustered network systems. IEEE Transactions on Control of Network Systems, 7(4):1736–1745, 2020
Nikitin, D., Canudas-de-Wit, C., and Frasca, P. A continuation method for large-scale modeling and control: from ODEs to PDE, a round trip. IEEE Trans. on Automatic Control, 67 (10): 5118-5133, 2021
Pratap, U., Canudas-De-Wit, C., and Garin, F. (2022). “Where, When and How people move in large-scale urban networks: the Grenoble saga”, submitted to Transportation Research Part C: Emerging Technologies.
Ruiz, L., Chamon, L.F., and Ribeiro, A.. Graphon signal processing. IEEE Trans. on Signal Processing, 69:4961–4976, 2021
D. Shah and C. Lee. Reducing crowdsourcing to graphon estimation, statistically. In International Conference on Artificial Intelligence and Statistics, pages 1741–1750, 2018.
Sosoe, K. “Modeling of multimodal transportation systems of large networks”, PhD thesis. Univ. Paris Est, 2017.
Pasquale, C., Siri, E., Sacone, S., and SIri, S. (2021). “ A discrete-time model for large scale multimodal transport networks” 16th IFAC symposium on Control in Transportation Systems (CTS 2021), Lille, France
Taia Alaoui, F., Fourati,H., Kibangou, A., Robu, B., and Vuillerme, N. Urban transportation mode detection from inertial and barometric data in pedestrian mobility. IEEE Sensors Journal, 22(6):4772–4780, 2022.

Contexte de travail

The Gipsa-lab is a joint research laboratory of the CNRS, Grenoble-INP -UGA and the University of Grenoble Alpes. It is under agreement with Inria and the Observatory of Sciences of the Universe of Grenoble. He conducts theoretical and applied research on AUTOMATICS, SIGNAL, IMAGES, SPEECH, COGNITION, ROBOTICS and LEARNING.
Multidisciplinary and at the interface between the human, the physical and digital worlds, our research is confronted with measurements, data, observations from physical, physiological and cognitive systems. They focus on the design of methodologies and algorithms for processing and extracting information, decisions, actions and communications that are viable, efficient and compatible with physical and human reality. Our work is based on mathematical and computer theories for the development of models and algorithms, validated by hardware and software implementations.
By relying on its platforms and partnerships, Gipsa-lab maintains a constant link with applications in a wide variety of fields: health, environment, energy, geophysics, embedded systems, mechatronics, processes and industrial systems, telecommunications, networks, transport and vehicles, operational safety and security, human-computer interaction, linguistic engineering, physiology and biomechanics, etc.
Due to the nature of its research, Gipsa-lab is in direct and constant contact with the economic environment and society.
Its potential as teacher-researchers and researchers is invested in training at the level of universities and engineering schools on the Grenoble site (Grenoble Alpes University).
Gipsa-lab develops its research through 16 teams or themes organized into 4 divisions:
• Automatic and Diagnosis (PAD)
• Data Science (PSD)
• Speech and Cognition (PPC)
• Geometries, Learning, Information and Algorithms (GAIA).
The staff supporting research (38 engineers and technicians) is distributed in the common services distributed within 2 divisions:
• The Administrative and Financial Pole
• The Technical Pole
Gipsa-lab has around 150 permanent staff, including 70 teacher-researchers and 41 researchers. It also welcomes guest researchers and post-docs.
Gipsa-lab supervises nearly 150 theses, including around 50 new ones each year. All the theses carried out in the laboratory are financed and supervised by teacher-researchers and researchers, including 50 holders of an HDR.
Finally, around sixty Master's trainees come each spring to swell the ranks of the laboratory.