“Predictive capabilities create entirely new, recurring revenue streams”
3 Questions for… Stefan Hassels, Head of Product, Compredict
Stefan Hassels speaks in Munich on the topic "Turning SDVs into a Business: How Virtual Sensors Cut Costs & Drive Revenue".
Compredict
In the run-up to the 6th International Conference Automotive Software Strategies, we asked Stefan Hassels, Head of Product, Compredict, three questions. The event will take place on 21 and 22 May 2025 in Munich, and Stefan Hassels is one of the speakers.
Before joining Compredict, he was employed at Audi. There he led product development processes and worked as Digital Strategist & Senior Business Designer. The focus of his presentation in the Bavarian capital will be on Virtual Sensors as an essential, business-critical layer in the SDV stack to reduce costs and enable revenue-generating services.
ADT: How do Virtual Sensors shift the economics of SDV feature development?
Hassels: Virtual Sensors fundamentally reshape the economics of SDV feature development by enabling cost savings at scale, unlocking new software-defined capabilities, and maximising aftersales potential. They deliver a dual benefit: reducing hardware dependency while enhancing the user experience. For example, by replacing a physical ride height sensor—traditionally used for automatic headlamp levelling—with a Virtual Sensor, OEMs can save €7-10 per unit when considering the sensor, wiring harness, and connectors.
At the same time, these Virtual Sensors elevate the vehicle’s intelligence, enabling services like predictive maintenance and personalised service offers. Just the enablement of tyre and brake wear monitoring can generate over €180 in additional revenue per vehicle across its lifetime. These aren’t incremental improvements—they represent a fundamental shift in vehicle economics. This is why Virtual Sensors are not just a technological enhancement, but a business-critical layer for every OEM aiming to build profitable, future-ready platforms.
What role do predictive models play in delivering data-driven, revenue-generating services?
Predictive models are at the core of COMPREDICT’s Virtual Sensors and are essential for OEMs and Tier 1s to remain competitive in an AI-driven, supply chain-sensitive industry. As customer expectations shift towards intelligent, data-informed experiences—particularly around vehicle health, safety, and performance—predictive models become indispensable. Just as modern businesses rely on data-driven decision-making, vehicle manufacturers must build customer-centric products powered by data-driven insights / algorithms.
These predictive algorithms do more than provide additional information. As mentioned in the earlier example with tyre and brake wear, predictive capabilities don’t just support the driver—they create entirely new, recurring revenue streams. In short, predictive models are the engine that transforms raw vehicle data into commercial value, making data a true competitive and financial asset.
How are your Virtual Sensors integrated into broader software stacks within the vehicle?
COMPREDICT’s Virtual Sensors are engineered for seamless integration into modern vehicle architectures. For example, they can be embedded directly on ECUs or even within the infotainment system—such as those running Android Automotive OS. This in-vehicle edge deployment ensures real-time performance, data privacy, and independence from external connectivity.
Our integration approach is modular and standards-aligned, ensuring that Virtual Sensors fit naturally into OEMs’ existing software stacks without requiring major architectural overhauls. This flexibility allows manufacturers to scale features quickly and cost-effectively across platforms—turning vehicle data into dynamic functionality.
