different software statuses more easily. The manual integration effort decreases, and, at the same time, the recurring validation tasks are executed in a more organized way. Increased efficiency in testing and troubleshooting The practical benefit lies in the precise reproducibility, which enables running critical scenarios multiple times, comparing software statuses with each other and pinpointing faults under stable conditions. As a result, test cycles get shorter and the validity of results increases. SensFlow synchronizes the recorded data of various sensor and bus systems This approach is especially effective with complex perception and localization functions. Tests become more independent from prototypes, weather conditions or rarely occurring traffic situations. At the same time, the real-world data basis necessary for a sound function evaluation is maintained. Integration into existing toolchains SensFlow was designed for seamless integration into already existing development and test environments. Combined with the CarMaker product family, the Image: © IPG Automotive replay approach can be extended to include closed-loop validation. This enables development teams to merge real measurement data and virtual scenarios into a continuous validation strategy. Raw data from camera, radar and lidar sensors as well as vehicle bus data via CAN or Ethernet are injected into ECUs. Existing HIL environments can hence be extended by adding real-world driving scenarios without having to rebuild the test infrastructure. In practice, SensFlow supports both real-world test drives and simulation: Real measurement data provide the input basis, replay enables reproducible use and virtual tests allow for a systematic expansion. This results in a scalable validation process that supports reproducible tests across the entire development process. Conclusion: Real-world data become a reproducible test basis Reproducibility is a prerequisite to validate automated driving functions. SensFlow enables the deterministic reproduction of real sensor and vehicle bus data, the test of software statuses under similar conditions and the targeted analysis of faults. Therefore, sensor data replay becomes a reliable building block for modern ADAS and AD validation. Development teams can use real edge cases reproducibly in the lab, shorten iteration cycles and evaluate software changes under comparable conditions. https://www.ipg-automotive.com/ solutions/product-portfolio/sensflow • finexus® – Nonwovens for lightweight and sustainable molded parts • Absorber nonwovens for acoustical and thermal insulation • Upholstery nonwovens for foam replacement • Efficient media for air and liquid filtration Excellence in Mobility sandler-group.com Filtration Sound Insulation Thermal Insulation Upholstery
RkJQdWJsaXNoZXIy MjUzMzQ=