Virtual sensors that utilize vehicle data reduce costs and create value

The transition to SDVs will enable vehicles to be equipped with high-performance processing capabilities.
By analyzing vehicle data with this processing power, it becomes possible to detect various vehicle conditions (vehicle status).
Virtual sensors are a new concept that uses software rather than physical sensors to perform sensing.

●Cost reduction
Physical sensor reduction: e.g., replacing headlight leveling sensors with virtual sensors
Reduce R&D costs: e.g., streamlining vehicle load testing

●New value creation
Visualization of consumables: For example, visualize the wear of tires and brake pads to provide timely maintenance services.
Update: Example: Realize new sensing with just a software update

A mobility data science company leading the SDV market

Virtual sensors created by combining automotive engineering and data science
Compredict: Founded in 2016 as a spin-off company from the Technical University of Darmstadt in Germany.
A software company that combines data science and automotive expertise to develop algorithms for virtual sensors. European OEMs plan to adopt Compredict's virtual sensors for tire and brake wear on 10 million vehicles by 2030.

New value created by software. Virtual sensors made possible by SDV and data analysis

The automotive industry is seeing advances in SDV architecture and integrated ECUs, and software-based value creation is gaining attention. This virtual sensor is a solution that captures this trend. The virtual sensor was created from knowledge gained by combining automotive engineering and data science.
The base model is provided by learning from the large amount of vehicle information collected by Compredict. The differences between the base model and each vehicle are also calibrated using experimental data from Compredict and the OEM, allowing for development in a short period of time.

Performance verification is already being conducted daily with multiple OEMs, and some OEMs have already decided to adopt the virtual sensor for commercial use, demonstrating their confidence in its performance.

SDV x Data Utilization: Cost Optimization and New Value Creation

Cost reduction

[Reducing physical sensors]

• headlight leveling sensor
  Automatic headlight levelling will be regulated by UN R48 and will become a mandatory feature from 2027.
  Compredict's headlight leveling virtual sensor is UN R48 compliant and can replace the hardware sensor.
  Replacing physical sensors with virtual sensors leads to cost reduction.

[Improving R&D efficiency]

• Wheel Force Transducer
  By replacing tests that require large-scale measuring equipment with virtual sensors, use cases where data could not be collected before (such as data on public roads) become possible, improving the efficiency of vehicle testing.
  Streamline R&D evaluation and experimentation.

New value creation

[Visualization of consumables]

• Tires, brake pads
  This visualizes the wear condition of tires and brake pads, which was previously impossible to sense constantly.
  This allows for timely provision of maintenance services.

【update】

• Adding a new sensing service
  After the vehicle is delivered to the end user, new sensing functions can be added simply by updating the software without installing any hardware. Service updates are also possible.
  Adding new features through software updates
  Generate revenue by developing aftermarket services using sensing

[Real-time vehicle weight monitoring]

• Preventing overloading
  This will enable enhanced safety measures and improved ride comfort, such as notifying drivers of overloading and adjusting autonomous driving brake control based on vehicle weight.

Compredict's Activity Topics

FAQ

Inquiry