Focus on engine management: innovations for the drive

The integration of electric drivetrain functions in a single component is one of the mega trends in drivetrain development.
Solutions are being presented worldwide and initial approaches are being brought into series production, with the Asian market leading the way.
The potential of highly integrated "X-in-1" solutions is clear: cost savings through reduced hardware and the use of synergies in the software. On the other hand, the individual components are more complex and repair costs are potentially high if a highly integrated component fails.
Using the "Powerbox" demonstrator as an example, IAV shows how the aforementioned potentials can be tapped and challenges addressed with intelligent solutions. The high level of integration also opens up new possibilities for functions within the electric drive that go beyond the state of the art.

Direct cooling of the winding using waveguides is revolutionising the thermal management of electric motors.
The potential for three times higher continuous current densities enables more compact, lighter drives with higher efficiency - a game changer for e-mobility.
Hyperdrives has set itself the goal of industrialising this technology on the basis of existing automotive production facilities.

Our company relies on advanced and innovative applications of artificial intelligence (AI) to optimize our services in various areas of the automotive industry.
In particular, we use AI to detect faults in electronic control units (ECUs). By using various machine learning methods and advanced algorithms, we identify and classify anomalies and thus offer quality assurance at the highest level.
We utilize a real time detection transformer trained with massive data as deep learning backbone. Our application enables technical experts to quickly assess the model predictions, modify, add or remove findings and analyze complete PCBAs at an utmost high standard. Moreover, we continuously generate valuable meta data by gathering the experts knowledge via lazy learning and improve the odels performance with every single use case.

The current challenges facing motor control systems due to legislation and the transformation of drive technology

Modern automotive software development is increasingly complex, requiring millions of lines of code and strict regulatory compliance. Traditional methods struggle with fast validation, cost efficiency, and quality demands. This presentation highlights how AI technologies – Generative AI, intelligent agents, and automated validation – can accelerate development and validation.

A key enabler of AI-driven validation is TestSphere, a modular ecosystem that integrates AI-powered au-tomation into software testing and verification. TestSphere leverages Retrieval-Augmented Generation (RAG) for efficient knowledge management, AI-driven test case generation for scalable validation, and Sherlock Trace, an intelligent solution for analyzing traces from SiL, HiL, and in-vehicle tests. Sherlock Trace applies pattern recognition and anomaly detection to extract meaningful insights from large trace datasets, significantly reducing the time required for debugging and compliance verification. The system seamlessly connects with established toolchains, reducing manual effort while increasing test coverage and accuracy. By incorporating AI-powered copilots for test design and requirement analysis, TestSphere enables up to 10x faster test case generation while maintaining compliance and quality. Additionally, cloud-based virtual validation environments allow developers to execute large-scale Software-in-the-Loop (SiL) tests efficiently, significantly reducing hardware dependencies and validation costs.

This session explores real-world cases where AI-driven validation significantly reduced development time and cost. It highlights best practices for integrating AI into automotive software engineering while ensuring security, scalability, and compliance. The discussion fosters idea exchange on AI’s transformative role in automated software validation and verification.

The transition of the automotive industry to the development of Software-Defined Vehicles (SDV) offers great potential for sustainable innovation - especially and including in engine control. At the same time, it presents us with major challenges in terms of complexity in managing new technologies.
This presentation presents a new perspective on the use of "golden data" for predictive maintenance and predictive diagnostics to both detect and fix problems in engine control and improve efficiency and performance through new paradigms introduced with SDV.

Leveraging 40 years of technological leadership, IAV developed a unified software architecture for all powertrains.
The approach features a powertrain abstraction interface that separates specific vehicle functions from generic components.
For SDVs, this enables decoupling of software and hardware, efficient reuse of legacy code, reduction of complexity, and straightforward integration of third-party components, increasing energy efficiency and driving comfort.
Our presentation focuses on a centralized torque path of an xHEV, including drive strategy, hybrid operation strategy and shift strategy.

Global megatrends such as digitalisation and the associated disruptive technologies, particularly in the context of artificial intelligence, are leading to new requirements for the vehicle drive system. The resulting challenges in development show that previous development methods are no longer able to cope with the degree of networking across the vehicle and the complexity. The question therefore arises as to how development methods can be further developed in order to exploit the opportunities arising from the digitalisation of the drive system and development methodology. The component-orientated development of the past is being further developed into a target image of the software-defined vehicle. This also means that drive development is transforming from drive module development to highly networked drive system development in the context of an overall vehicle architecture.