Fuel cells are considered a key technology for the energy transition, as hydrogen can serve as a sustainable energy carrier when produced from renewable sources. However, water generated during operation can freeze under sub-zero conditions, leading to start-up failure and accelerated degradation. Reliable freeze-start capability is therefore a critical requirement for fuel cell electric vehicles (FCEVs). In this study, the different phases of a freeze-start strategy, including freeze preparation, stack cooling, and freeze start with internal heat-up, were experimentally investigated on an automotive-sized 10-cell short stack equipped with an adapted cooling circuit to reproduce vehicle-relevant thermal boundary conditions. The influence of freeze-preparation parameters and freeze-start operating conditions was analyzed. The results show that the extent of membrane drying is the dominant factor determining freeze-start capability. Increasing the high-frequency resistance (HFR) threshold after drying from 0.20 to 0.25 to 0.60 Ω·cm²/cell improved the minimum achievable start temperature from approximately − 10 °C to − 18 °C without negative cell voltages. Furthermore, lower stack voltages during potentiostatic heat-up increased the heating rate from 0.26 to 0.32 K s⁻¹ while simultaneously reducing the total water production from 10.3 g to 8.8 g. Variations in cathode airflow had only a minor influence on freeze-start performance. Overall, the study demonstrates that combining an aggressive but controlled freeze preparation with low-voltage potentiostatic heat-up provides the most robust and efficient freeze-start strategy for automotive PEM fuel cell stacks under severe winter conditions.
Driver behaviour in the dilemma zone critically influences red-light violations and intersection safety. This study presents a scalable Vision-AI and deep learning framework modelling driver decisions based on vehicle type, speed, acceleration, lane occupation, and signal distance. Validated at a signalised intersection in Thessaloniki, Greece, UAV-collected data yielded over 11,000 vehicle trajectories annotated with stop/go decisions during the yellow phase. A multi-layer perceptron classifier trained on this data reached 92.33
The rapid evolution of autonomous driving technologies is deeply reliant on high-quality datasets and robust data management solutions that cannot be achieved by a single actor working in isolation. This paper examines the growing role of datasets and data spaces in the mobility domain, providing a conceptual analysis of current approaches and identifying key challenges and emerging trends that inform the design of an autonomous vehicle data space framework. While peer-to-peer solutions for data sharing already exist, the concept of data spaces can significantly scale these approaches, creating a self-improving feedback loop that enhances the performance of AI-based driving models by matching the demand for data with the data offered by providers. Specifically, the paper addresses the challenge of acquiring intra-vehicular data to enhance assisted driving functionalities and proposes a methodology for building a data space specialized for automated driving applications. Finally, the paper explores representative use cases, including a Level 4 automated shuttle, TalTech iseAuto, Estonia, and an all-terrain vehicle provided by CERTH, Greece, by conceptualizing data sharing across different domains, where data providers and consumers exchange information within a data space framework, driving continuous improvement in model performance.
Urban transport systems face mounting pressures from climate change, rising energy prices, and ageing assets. Meeting these challenges sustainably requires extending the useful life of infrastructure, improving energy efficiency, and reducing waste across the whole lifecycle. Circular economy principles offer a systemic approach: design for longevity, maintainability, and reuse, while favouring interventions that avoid resource-intensive renewals. Digitalisation is a key enabler of this shift, as Artificial Intelligence-assisted analytics, edge sensing, and Digital Twins make hidden conditions visible and reduce diagnostic costs. This paper presents three European pilot cases from the Interreg CE4CE project to demonstrate these capabilities. In Leipzig, a layered infrastructure monitoring system enabled the detection of eight critical overhead line faults and three previously undetected broken rails during regular passenger service. In Bergamo, high-frequency onboard sensing and Energy Flow Simulations optimized energy recuperation and identified pantograph-infrastructure “hot spots.” In Gdynia, a Digital Twin environment was used to simulate over 110 electrification scenarios, optimizing fleet size and infrastructure investment needs. The results confirm that integrating digital monitoring and simulation tools into public transport operations facilitates a shift from reactive to predictive management, directly supporting resource efficiency, asset longevity, and the transition toward a circular urban mobility system.
Abstract Sustainable forestry is fundamental for the supply of raw materials for sustainable products such as wooden houses, packaging, and biofuels. Efficient transport is vital for sustainable forestry, as road transports of roundwood from the forest to recipient industries account for 50% of CO2 emissions and 40% of the costs for harvesting and transport. Transport of roundwood is characterised by deliveries from many sources to few industries. As wood is heavy, weight generally restricts load capacity, and an increase in gross vehicle weight of log trucks to increase payloads may reduce costs and emissions. Limitations in road bearing capacity might necessitate detours and longer driving distances if the full load capacity of heavier trucks is to be used. To comply with road regulations, heavier trucks must have more axles, which increases investment cost, rolling resistance and tare weight, and thereby transport costs and emissions per vehicle-km. If larger trucks are used on roads with low bearing capacity, their payloads are lower than for standard 7-axle trucks. We studied how truck fleets with a mix of truck sizes can be used to reduce costs and the influence of the road bearing capacity using a discrete event simulation model developed for a case in Sweden. A base scenario with a fleet of 7-axle vehicles was compared to nine alternative fleet compositions using three performance indicators: cost per m3, distance driven per m3, and time per m3. These fleets were operated on three road networks: A, the current mix of 64-tonne and 74-tonne roads; B, major roads upgraded to 74 tonnes; and C, all roads upgraded to 74 tonnes. Results show that adapting the fleet composition to the available road network is beneficial from work time, fuel consumption and cost perspectives. Although it is beneficial to strive for heavier vehicles, the road network restricts the number that are useful. When heavier vehicles are used on 64-tonne roads there will be losses in terms of costs, fuel consumption, and payload. For the current road network, a fleet of 7- and 9-axle vehicles will be preferable if transports can be planned. Upgrading roads from 64 to 74 tonne bearing capacity will decrease work time, distance travelled, and costs per m3 transported. When considering societal costs for road upgrades, the benefit will probably be smaller for less used roads. As the design of a truck fleet is complex, simulation is a useful tool to assist hauliers in evaluating fleet compositions.
