Traffic safety is the fundamental criterion for vehicular environments and many artificial intelligence-based systems like self-driving cars. There are places, e.g., intersections and shared spaces, in the urban environment with high risks where vehicles and vulnerable road users (VRUs) such as pedestrians and cyclists directly interact with each other. By advancing starte-of-the-art artificial intelligence methodologies, this project VeVuSafety aims to build a privacy-aware deep learning framework to learn road users’ behaviour in various mixed traffic situations for the safety between vehicles and VRUs. VeVuSafety proposes a 3D environment model based on 3D point cloud for privacy protection — private information like license plates and face is anonymized. Then, within this environment model, an end-to-end deep learning framework using camera data will be built for multimodal trajectory prediction, anomaly detection, and potential risk classification based on deep generative models such as Variational Auto-Encoder. Additionally, an active privacy mechanism will also be adopted by application of the differential privacy mechanism to help the deep learning models prevent model-inversion attack. Moreover, the framework’s generalizability will be investigated by exploring the Normalizing Flows approach for domain adaption. The framework’s performance will be validated at different intersections and shared spaces using real-world traffic data. Besides road user safety and privacy, VeVuSafety can help traffic engineers and city planners to better estimate the design of traffic facilities in order to achieve a road-user-friendly urban traffic environment. Furthermore, the success of VeVuSafety will enhance the fellow’s scientific knowledge and project management skills to become an artificial intelligence expert for traffic safety and Intelligent Transportation Systems.

The increasing technologization of society, combined with the transition to renewable energy sources, requires materials and architectures that can handle and retrieve large amounts of data on short time scales and with minimal power consumption. As a result, piezoelectric materials have attracted significant research interest. However, lead-free BaTiO3-based piezoelectric materials currently cannot match the performance of traditional lead-based Pb(Zr,Ti)O3 materials. The MoBaFi project aims to address the challenge of enhancing the piezoelectric response and increasing the Curie temperature of lead-free BaTiO3-based materials. Overcoming these challenges is crucial for improving performance and advancing applications. The scientific approach of the MoBoFi project will focus on developing sophisticated compositional and microstructural modifications in BaTiO3-based films, along with an understanding of the relationship between microstructure-property via in situ characterizations. This project facilitates a two-way and transdisciplinary transfer of knowledge between (1) the applicant’s expertise in the chemical solution deposition (CSD) method and ex situ microstructural characterization, (2) the strong expertise of pulsed laser deposition (PLD) method and in situ ferroelectric characterization at the research group of University of Twente (Host, Supervisor Prof. Gertjan Koster), and (new) the applicant’s own trajectory in building expertise in in situ microstructural characterizations. A secondment will take place at two departments of Ghent University (Prof. Klaartje De Buysser and Prof. Stefaan Cottenier) for size-controlled nanocrystal synthesis and computational screening of ABO3 perovskites. This MSCA-PF will certainly contribute to training the applicant in new scientific and transferable skills, enhancing the career perspectives to become an independent and mature researcher in the EU in the near future.

Our healthcare system is under unsustainable strain owing, largely, to cardiovascular diseases and cancer. For both, imaging vasculature and flow precisely is paramount to reduce costs while improving diagnosis and treatment. Specifically, the focus is on the multiscale aspects of shear, vorticity, pressure and capillary bed (10-200 μm vessels) structure and mechanics. However, this requires an imaging depth of ~10 cm with a resolution of ~50μm. Furthermore, velocities often exceed 1m/s, which requires a frame rate of ~1000 fps. Clinical imaging modalities have so far been hindered by insufficient spatiotemporal resolution and there is thus a dire need for new techniques. Plane-wave ultrasound enhanced with contrast microbubbles outperforms all modalities in safety, cost, and speed, and is thus the ideal candidate to address this need. The strategy I propose in Super-FALCON harnesses the nonlinear dynamics of monodisperse microbubbles. In WP1, I will use deep learning and GPU-accelerated acoustic simulations to recover super-resolved (1/20th of the wavelength) bubble clouds. In WP2, I will create a new model for confined bubbles, and use them as nonlinear sensors for capillary imaging. In WP3, I will disentangle attenuation and scattering using (physics-informed) deep learning and correct for wave distortion. This is needed to apply the strategies from WP1 and 2 in deep tissue. Finally, in WP4, I will use automatic segmentation to integrate the fundamental results of WP1, 2 and 3 into a technology that I will scientifically assess on vascularized ex vivo livers. With Super-FALCON, my ambition is to generate a long-term impact both scientifically and societally. I will produce new fundamental knowledge about confined bubble dynamics, inhomogeneous ultrasound propagation, and deconvolution strategies as well as new experimental methods for flow imaging and characterization. In healthcare, Super-FALCON could initiate a paradigm shift towards patient-specific treatment.

Endovascular interventions are an established class of procedures within minimally invasive surgery (MIS). They enable the treatment of cardiovascular diseases through small incisions in the body by using flexible instruments. Conventionally, these instruments are manually operated, which restricts their precision, and limits their applicability. The magnetic actuation of instruments for endovascular interventions creates a novel and effective steering alternative. Even in deeply seated regions, magnetic flexible instruments provide clinicians with dexterity, while retaining minimal access. Thereby, they permit a range of advanced surgical tasks unattainable otherwise. Nevertheless, to be remotely actuated, such instruments rely on magnetic fields originating from outside the body. Thus, the aim of RAMSES is to develop and evaluate a clinic-ready robotic system capable of generating external fields during endovascular interventions. The RAMSES system will become an enabling technology for the clinical use of magnetic surgical instruments. It will truly revolutionize MIS, opening a new market for advanced diagnosis and treatment options. The RAMSES system will contain optimized electromagnetic actuators, located on robotic manipulators and powered by dedicated control software. The resulting versatile clinical framework will be applicable to a wide range of surgical instruments. This includes both commercially-available magnetic catheters as well as novel experimental designs. As a consequence, RAMSES will satisfy the needs of clinicians to further expand the effectiveness and availability of MIS techniques. It will provide an indispensable clinical tool for accurate and comprehensive surgical interventions in hard-to-reach locations within the human body. RAMSES aspires to turn magnetic actuation into a commercially successful technology. It involves strong industrial collaborations and a dedicated business development team in an ambitious quest to make that happen.