Bosch project uses AI to refine data collection for autonomous vehicle testing
Bosch Research and partners have presented results from the KIsSME project, which developed an AI system for smarter data logging during autonomous vehicle test drives.
Renningen, Germany – Bosch Research, along with its consortium partners, presented the outcomes of the KIsSME project on November 28. The three-year initiative, funded by the German Federal Ministry for Economic Affairs and Technology, focused on developing an AI-driven system to intelligently identify and record critical data during test drives of highly automated vehicles.
The project addresses a major challenge in the development of autonomous driving: the sheer volume of data generated. Test vehicles equipped with numerous sensors produce gigabytes of data per second, exceeding the capacity for conventional processing and analysis. The KIsSME project has yielded a software prototype that utilizes AI algorithms to evaluate traffic situations in real time, selectively storing only the data that offers significant value for improving vehicle performance.
"You could say the project has sharpened the senses of automated test vehicles," stated Daniel Schmidt, Bosch coordinator for the KIsSME project. "Intelligent data storage helps make the development process for safe automated vehicles quicker and more efficient." The captured data is used to train the vehicles' maneuver algorithms, enhancing their ability to navigate complex traffic scenarios.
In addition to Bosch, project participants included RA Consulting GmbH, Mindmotiv GmbH, LiangDao GmbH, the Fraunhofer Institute for High-Speed Dynamics (EMI), the Research Center for Information Technology (FZI), and the Karlsruhe Institute of Technology (KIT). The consortium was led by AVL Deutschland GmbH. Bosch intends to incorporate the project's findings into its ongoing development of safe automated vehicles and future driver assistance systems across its divisions.
The final results of the KIsSME project were showcased at the concluding event held at Bosch Research in Renningen, attended by all project partners. Artificial intelligence played a crucial role throughout the project, from image recognition and assessment of driving maneuvers to the anonymization of sensor data.