ValgrAI organised an informative event at the ETSINF (Escuela Técnico Superior de Ingeniería Informativa) that brought together the VI Meeting of the Joint Research Unit (UMI) of ValgrAI and a Conference on AI and Mobility within the framework of the Renault Group- UPV Chair.
The VI UMI Meeting was opened by Ana Cidad, Managing Director of ValgrAI, who highlighted the strategic role of the entity as a hub for talent, research and technology transfer in artificial intelligence. This was followed by Vicent Botti, Managing Director of ValgrAI, who reflected on digital twins, Smart cities and mobility.
Eva Vallada, researcher at ITI-UPV and ValgrAI, opened the session with a presentation on the optimisation of the management of emergency medical vehicles. Using mathematical models applied to the problem of location and relocation of ambulances, she showed how tools such as Hireves can substantially improve service times, especially when demand is high. The heuristics developed make it possible to anticipate coverage shortage scenarios and plan strategic relocations that increase the number of calls attended in less than 10 minutes.
Iván Arcos, Data Science researcher at the Pattern Recognition and Human Language Technology (PRHLT) research centre. Univeristat Politècnica de València, addressed the growing problem of sexism in social networks, focusing on the TikTok platform, characterised by its huge volume of ephemeral and rapidly disseminated content. Through the EXIST 2025 project, a system capable of automatically detecting and classifying sexist content in videos, images and text was presented. To do so, a multilingual dataset was compiled from more than 200 hashtags potentially associated with sexism, and specific tasks were defined such as the detection of sexism, the classification of the intention (direct or reported) and the categorisation into different forms (objectification, stereotypes, ideology, sexual and non-sexual violence).
Javier Palanca, researcher at VRAIN- UPV, presented a digital twin system applied to urban traffic management implemented on SPADE, the multi-agent systems platform (MAS) developed at his institute. The project digitally models 386 street sections in the city of Valencia. Each agent collects real-time data (such as traffic or weather) and allows simulating street closures, evaluating traffic redistribution strategies and improving the response to unexpected events.
Mercedes Landete, from the Operations Research Centre of the Miguel Hernández University, discussed emerging trends in route calculation using artificial intelligence, focusing on the last mile, drone logistics and sustainable delivery. She highlighted the importance of integrating multiple variables (type of vehicle, weight, congestion, topography, etc.) and advocated the use of machine learning algorithms (such as random forest or gradient boosting) to dynamically adapt to complex and volatile urban environments.
The second part of the event, the AI and Mobility Day, was introduced by Carlos Guindel (Renault Group, Head of ADAS/AD Fusion), Leticia Picado (Renault Spain, Software & Technology) and José Hernández-Orallo (ValgrAI-VRAIN Universitat Politècnica de València).
After them, ValgrAI researcher Carlos Monserrat (VRAIN-UPV) presented the progress of the European project HIDDEN (Advanced AI-based collective awareness in autonomous driving). The aim of the project is to develop technologies that improve the collective perception of hidden obstacles in autonomous driving, through distributed sensors, trajectory prediction and processing with language models. This shared awareness between vehicles makes it possible to anticipate risk situations and alert the driver in a comprehensible way, even when visibility is reduced or the danger is outside the field of vision.
Vicente Julián, also a researcher at ValgrAI and VRAIN-UPV, focused his presentation on the SimFleet simulator, a tool based on multi-agent simulation that allows the study of urban fleet management strategies (taxis, couriers, car-sharing, etc.). Its latest evolution incorporates agents based on language models (LLM), capable of reproducing complex human decisions with memory, adaptive reasoning and natural interaction capacity. This architecture opens up new possibilities for urban behaviour modelling and transport policy planning.
The final round table, moderated by José Hernández Orallo, was attended by Carlos Monserrat, Carlos Guindel and Vicent Botti. The debate revolved around the ethical, regulatory and technical implications of the implementation of AI in autonomous transport systems, as well as the role that public administrations and the business sector should play in its safe deployment.
The speakers agreed that one of the main challenges of autonomous driving is not only technological, but also regulatory and social. Indeed, Carlos Guidel pointed out that there is still a gap between current levels of automation and the point at which the driver would no longer be necessary. To reach that level, it is essential to ensure that the system works correctly in all situations, without errors or ambiguities.
Vicente Julián also emphasised that, although the technology is quite mature, the real challenge lies in its acceptance by society and in the definition of new regulations. This development represents a paradigm shift in mobility that will affect entire sectors of work and the way in which transport is understood.
Vicent Botti pointed out that assisted driving already works well in controlled environments, but when it is introduced in a city, where humans and non-automated vehicles coexist, decisions become more complicated. In his view, the key is to differentiate between an AI that assists in decision-making and a fully autonomous AI, similar to the autopilot in aviation that always needs human supervision. Julian agreed, adding that the real challenge is to make decisions in real time. For him, the immediate future lies in hybrid environments, where the coexistence between autonomous systems and humans generates the most complex situations. Lastly, the role of vehicle owners was addressed, especially in Europe, where regulation is pioneering and privacy is being safeguarded. The need for the use of data to always be based on user consent was highlighted, and it was stressed that AI should help improve services without imposing a forced change of habits.
