1. Geely Automotive Group Introduction

Basic information about Geely Automotive Group, including the company's vision, mission, and values, as well as its industrial layout, vehicle development, vehicle validation, and vehicle manufacturing global footprint.

2. Intelligent Vehicle Evolution Path

Industry drivers for the continuous evolution of intelligent vehicles, and the future direction and goals of intelligent vehicle evolution.

3. Geely Automotive Electronic and Electrical Architecture Roadmap

Geely Automotive's electronic and electrical architecture roadmap in the development of intelligent vehicles, including an introduction to the latest GEEA 3.0 architecture launched in 2024, and the vision for the next-generation EEA 4.0 architecture.

4. Continuously Evolving Operating System Platform - GOS

Introduction to Geely's self-developed operating system platform (GOS), its key technologies, agile iteration philosophy, and the creation of a closed-loop vehicle ecosystem.

5. Outlook and Collaboration

Future industry outlook.

The car is no longer a closed, isolated system. As software becomes the core of vehicle functionality and customer experience, development must span two fundamentally different worlds: onboard (embedded, reliable) and offboard (cloud-based, flexible) systems. In this keynote, we explore how these two domains can learn from each other to create a unified, high-performance software delivery and operations model.

We will examine what cloud-native software practices like DevOps, Continuous Deployment, and Observability-Driven Development (ODD) can offer to embedded automotive environments — and, just as importantly, what the rigor, validation discipline, and reliability mindset from onboard software can teach the cloud world. Topics include OTA updates, End-to-End Software Delivery Journey and traceability.

The audience will leave with a clear understanding of how cross-pollination of methods and mindsets between onboard and offboard development can drive innovation, improve quality, and reduce time to market — without compromising reliability or flexibility. The next generation of cars will be integral part of a software defined ecosystem. It will rely on a shared engineering culture that combines the strengths of both worlds to build smarter, reliable vehicles that adapt to the customers expectations and needs. 

The automotive industry is undergoing a significant paradigm shift driven by the rapid advancements in China, which transformed from a follower to a leader in speed and innovation. Traditional automotive players in markets like Europe and North America have lost their technological edge. The shift from mechanical machines to electronic devices is evident, yet many established companies fall behind, moving too cautious and inward focused. To remain relevant in what is proclaimed to be the software defined vehicle era, the auto industry is finally forced to adapt to Consumer Electronic like time-to-market and empathetic user-centric experiences. 
Harman, with the CE juggernaut Samsung as mothership, emphasizes starting with user stories rather than requirements documents, focusing on swift deployment and continuous feature releases rather than the traditional fire & forget mentality in Automotive. Cross-domain collaboration, unified self-describing APIs, and a holistic understanding of user expectations have become essential. 
As the industry faces crisis, this is the chance to pivot: from project-based traditions to productized solutions, fostering stronger cross-domain collaborations and thus maintaining relevance in this digital age. This journey, though challenging, is crucial for survival. Delivering products that inspire and engage consumers is the AllSpark of the next generation of the Automotive industry.
With this spirit we will share a deep insight into our journey from the initial idea in 2021 to the holistic in-cabin experience product portfolio as presented at CES 2025, and beyond.

The automotive market is undergoing a clear shift from hardware-centric vehicles to software-defined vehicles (SDVs). As innovation at greater speed has become the new norm, household-name businesses are faced with the challenge of preserving their competitive edge, once built on manufacturing and supply chain expertise. At the same time, emerging software-based businesses require experience in developing and industrializing automated driving systems, to launch their products at scale.

This challenge is particularly evident in the realm of automated driving, where the ability to scale is crucial for achieving a meaningful impact.

Complementary partnerships and business models that bring the most value to customers, have the power to tackle such market challenges, push innovation forward, and alleviate societal- and market-related strains.

This presentation will provide a concrete example of such a complementary partnership – between Continental and the self-driving vehicle technology company Aurora, with the aim to deliver a commercially scalable fully autonomous trucking system.  Continental brings systems competence, knowledge of hardware, and supply chain expertise, while Aurora complements it with its industry-leading autonomous technology. 

The partnership is based on the industry-first hardware-as-a-service business model, based on mileage driven, with a goal of delivering maximum uptime and value to customers. 

Such examples aim to confirm that, although the future of mobility is autonomous, it is highly dependent on partnerships, collaboration, and joint strategies.

To encourage a discussion and questions, the presentation will conclude with a thought-provoking question for the attendees.

A clear understanding of the driver's cognitive state is essential for improving driver monitoring technologies and smarter mobility, yet current systems fall short of this goal. This gap poses a serious challenge to driver safety and has been highlighted by regulatory bodies like Euro NCAP and NHTSA in the USA, which stress the need for more advanced solutions for monitoring drivers’ cognitive state.

To address this, CorrActions’ researchers are using advancements in neuroscience, artificial intelligence, and machine learning to develop new methods for real-time cognitive state assessment. Using patented neuroscience technology, brain activity is analyzed through muscle micro-movements, leveraging data from existing in-cabin sensors like the steering wheel.

CorrActions’ passive and non-invasive approach, allows for accurate and direct detection of impairments such as alcohol intoxication, fatigue, and cognitive overload—not just interpreting outward signs but analyzing actual brain activity. It also plays a key role in supporting future advancements like autonomous driving systems and occupant wellness technologies, ensuring an optimal and trusted driving experience.

Making use of existing vehicle hardware & data - the technology meets and suppresses requirements set by standardization bodies like Euro NCAP’s recently released protocol and challenges the traditional belief that better safety requires more sensors, thus enhancing safety for mass adoption without increasing costs.

Vehicle autonomy remains a niche market primarily focused on robotaxis. However, north-star companies like Waymo have proven that with the right sensors and software, autonomy is complex but not insurmountable. The challenge is scale. 
At the other end of the market almost all automotive OEMs are stuck on level 2, eyes on/hands on, features. The sensor hardware and integration costs beyond Level 2 has led to low adoption. 
LiDAR, despite significant investment has failed to scale due to sensor cost and integration complexity. Radar and cameras are low cost and dominate the ADAS market but to date have lacked the perception performance required beyond level 2. 
We now have imaging or 4D radars delivering LiDAR like performance but with radar robustness. But imaging radar must learn from LiDARs failing and concurrently achieve the perception performance without losing the size, weight, power and cost advantages it has had to date. 

This needed a paradigm shift in radar design and was why Provizio has developed the software defined antenna, MIMSO, that extracts 20x more resolution from every radar IC. This allows high performance at a lower cost and unlocks ubiquitous, safe vehicle automation.

Bidirectional charging presents promising business opportunities by integrating vehicle batteries into home energy systems and the power grid. This technology leverages technical potentials such as increasing photovoltaic (PV) self-consumption, load shifting with dynamic electricity tariffs, peak shaving, and providing balancing power to stabilize the grid. These potentials translate into monetary savings and revenue opportunities. However, realizing these benefits requires the collaboration of various stakeholders, including end customers, energy suppliers, and aggregators. For bidirectional charging to achieve widespread adoption and fully realize its potential, it must be economically viable for all parties involved.

Several challenges must be addressed to make bidirectional charging economically feasible. Efficient voltage conversion between the power grid and the vehicle is crucial for most applications. Currently, inefficiencies in inverters under partial load and high auxiliary power consumption of control units hinder the effective and economical implementation of these applications. To address these issues, FEV is developing comprehensive solutions, including modular designs of voltage converters, optimized operating strategies for high utilization of voltage converters, adapted E/E architectures, and control unit optimizations.

In addition to these losses, the potential accelerated aging of batteries due to increased cycling is another barrier to the widespread adoption of bidirectional charging. FEV proposes optimal cycling strategies and smart charging techniques to mitigate cyclic and calendar aging as potential solutions.

This presentation will explore the potential use cases for bidirectional charging, the technical challenges and solutions mentioned, and discuss how and who can benefit financially from bidirectional charging.

The automotive industry is undergoing a transformative shift, driven by advancements in technology and the increasing demand for innovative, efficient, and sustainable solutions. At the forefront of this transformation is agentic AI, a powerful tool that is revolutionizing automotive research and development (R&D). This presentation will explore the multifaceted benefits of agentic AI in automotive R&D, highlighting its potential to accelerate innovation, enhance product quality, and reduce costs.

Agentic AI, with its ability to autonomously generate and process data, offers unprecedented opportunities for automotive manufacturers. By integrating agentic AI into the R&D process, companies can streamline complex tasks, such as design optimization, predictive maintenance, and autonomous driving systems development. This not only shortens the time-to-market for new vehicles but also ensures higher precision and reliability in product development.

Moreover, agentic AI facilitates a more agile and adaptive R&D environment. It enables real-time data analysis and decision-making, allowing researchers to quickly respond to emerging trends and challenges. This adaptability is crucial in an industry characterized by rapid technological advancements and shifting consumer preferences.

The presentation will also address the strategic implementation of agentic AI, emphasizing the importance of a value-centered approach. By focusing on key areas where AI can add the most value, automotive companies can maximize their return on investment and drive sustainable growth.

Customer expectations towards AI in automotive are driven by the evolution of AI experiences in the CE ecosystem. BMW’s vision is to make driving a holistic intelligent experience based on AI being an integral part of the entire vehicle. 

BMW believes 5 architectural paradigms will be key to deliver on this promise: 

• AI@Edge: Performance, reliability, and privacy requirements drive the need for onboard AI compute, which must keep up with CE innovation cycles.
• Consistency: A holistic intelligent experience from decentralized function development requires an overarching context model and function orchestration.
• Multimodality: To ensure the axiom “see what the customer sees”, a vehicle-spanning video architecture is key.
• Cross-domain interconnection: Remarkable customer value will be enabled by deeply integrating domain functionality, such as autonomous driving and digital cabin interfaces.
• Interchangeability: Flexibility to act on regional market and vendor dynamics make flexible AI architectures and versatile platform APIs paramount. 

In the near-term, OEMs should therefore focus on three success factors: 

• AI platforms: Cross-domain AI stacks to ensure seamless intelligent behavior in the vehicle.
• Context models: Vehicle and user data to differentiate features from a generic intelligence.
• Value creation cuts: Clear development focus on shaping and integrating AI experiences. 

In general, automotive AI architectures (cloud/edge) and value chains will profit from industry-wide patterns to allow for an efficient fit to outside partners and ecosystems. 

The automotive industry is at a crossroads: How can companies lower the Total Cost of Ownership (TCO) while driving innovation at full speed? Established Tier 1 suppliers now battle not only each other but also fast-moving Asian manufacturers and Electronics Manufacturing Services (EMS) providers, which already excel in hardware development and are in the process of catching up in automotive software development. 

Especially Western automakers face the tough challenge of cutting costs and accelerating at the same time. The Software-Defined Vehicle (SDV) approach promises both, but only a few industry players have the expertise to scale it effectively.

Meanwhile, China's automotive software ecosystem is setting a rapid pace, seemingly offering the perfect formula of efficiency and speed, complemented by exceptional versatility and a robust business mindset. 

But will regulatory barriers and geopolitical tensions disrupt this model’s global viability?

What are the essentials we need to focus on in this rapidly evolving landscape? What is required to get the best out of the promising solution “SDV”? 

This presentation provides answers on how to adapt to these new global realities. In focus are the emerging alliances of Western and Eastern companies, how responsibilities in the supply chain are shifting, what this means for business processes and monetization models and the continuing need for a change in mindset.