The Transport Research Arena (TRA) Conference 2026 in Budapest once again brought together leading voices from research, industry, and civil society to exchange perspectives on the future of mobility. Under this year’s theme, “Re-Generation in Transport”, discussions focused on how transport systems can respond to rapidly evolving environmental, technological, and societal demands.
Within this framework, the EFFEREST project contributed its latest developments in intelligent and adaptive energy management for electric vehicles, highlighting approaches that place user requirements, system efficiency, and predictive intelligence at the centre of vehicle development.
Nikolai Ebinger and Alex Kospach presented EFFEREST’s poster on integrating real user expectations into advanced energy-management strategies.
A central challenge in electric mobility lies in balancing several often-competing objectives: thermal comfort, energy efficiency, battery performance, and driving range. Rather than treating these aspects independently, EFFEREST investigates how they can be coordinated through integrated, adaptive control concepts.
The presented work builds on structured user interviews, through which mobility needs, preferences, and expectations were analysed and translated into functional system requirements. These requirements form the basis for more responsive and personalized vehicle behaviour.
The showcased approach combines several complementary concepts:
- Holistic User-centric Control (HUC) to dynamically balance occupant comfort and energy efficiency
- Model Predictive Control (MPC) to optimize energy flows by anticipating future driving situations and individual user preferences
- Integrated thermal and powertrain management, replacing isolated subsystem optimization with coordinated system-level control
- Adaptive Digital Twins (ADTs) enabling personalized eco-features that help reduce range anxiety while preserving comfort and efficiency
EFFEREST was also represented in the presentation “Adaptive Surrogate Model for Real-time Condition and Health Monitoring of Powertrain Components”, delivered by Manuel Van Rensbergen (VUB) and Alex Kospach.
This work addresses another important dimension of next-generation vehicle intelligence: the ability to monitor component condition and health in real time.
The presented approach introduces a deep-learning-based surrogate model for a traction inverter, derived from a high-fidelity multi-physical simulation model. By combining LSTM neural networks, state segregation methods, and physics-informed training strategies, the model enables highly accurate real-time predictions while maintaining computational efficiency.
Achieving prediction accuracies above 98% (R² > 98%), the methodology demonstrates strong potential for deployment within Adaptive Digital Twins for automotive powertrain systems.
Beyond accurate modeling, this capability opens pathways toward:
- real-time system monitoring
- predictive maintenance strategies
- intelligent, data-driven vehicle operation
- scalable digital-twin applications for automotive systems
Participation at TRA 2026 provided an important opportunity to engage with the broader mobility research community and contribute to ongoing discussions on adaptive, user-centric, and predictive technologies.
By combining advanced control strategies, artificial intelligence, and digital-twin methodologies, EFFEREST continues to explore new pathways toward electric vehicles that are not only more energy-efficient, but also more personalized, resilient, and responsive to real user needs.
