Technology Artificial intelligence (AI) and internet of things IOT  network animation concept. Big data flow analysis, deep learning modern computer technologies network connectivity.
Continental in the lead

‘KI Wissen’ Ends With Final Event in Frankfurt 

PPP project enhancing AI knowledge

Continental leads large PPP project enhancing KI Wissen (translated AI knowledge) to contribute to making autonomous driving safer and more reliable.

March 22, 2024 - KI Wissen  is the fourth and last Public Private Partnership (PPP) project of the project cluster KI Familie (translated AI Family) which is part of the VDA Leitinitiative (lead initiative by the German Association of the Automotive Industry), an initiative that sets clear standards for safe autonomous driving and plays a crucial role in identifying and jointly developing important topics and action areas for the progress of autonomous vehicle functions. To gain further valuable insights into the other three projects within KI Familie, please find more information at the end of the page.

KI Wissen Project Team

By integrating rule-based knowledge with artificial intelligence systems and refining appropriate methodologies, we are paving the way for the future of mobility together with our partners within KI Wissen. Enhancing our automotive software functions with AI-based algorithms is crucial e.g., to ensure the reliability and safety of autonomous driving.


Gilles Mabire

CTO Continental Automotive / Head of Software and Central Technologies (SCT)

How KI Wissen started:

Beginning in January 2021 KI Wissen aim was to explore how established knowledge like traffic regulations, mathematical-physical principles or social norms can be seamlessly integrated into AI systems. Under the lead of Jörg Dietrich, Head of AI for R&D Excellence at Continental, this hybrid approach, combining data-driven techniques with knowledge-based methods, seeks to redefine the foundation for training and validating AI functions. With a budget exceeding 25 million euros and supported by the German Federal Ministry for Economic Affairs and Climate Action, 15 project partners came together to address key hurdles on the path to autonomous driving: generalizing AI to phenomena with limited data, enhancing stability against data disturbances, ensuring data efficiency, validating and securing AI-assisted functions, and ultimately elevating functional quality.

KI Wissen Gilles Mabire
KI Wissen Jörg Dietrich
KI Wissen Simon Heinz

Set-up and results of KI Wissen:

Simon Heinz, Project Owner of AI Enterprise Applications at Continental, who took over the coordinating role from Jörg Dietrich in early 2023, outlines the benefits for Continental Automotive of taking on a bigger role in the KI Wissen project: “Building up expertise in and distributing the knowledge about hybrid AI throughout Continental early will be very helpful for many future external and internal AI-relevant projects. It will help us to expand our network in the field of automotive AI and attract new talent as well as keeping them in our company.”


To make reaching this goal more feasible, KI Wissen divided the overarching main task into four essential sub-projects:  

  • Knowledge Integration was to develop methods to integrate domain knowledge like physical laws or world knowledge into training processes for existing and novel architecture and AI components, laying the groundwork to enhance data efficiency, generalization, and safety.  
  • Knowledge Extraction focused on developing concepts and measures for extracting knowledge from AI processes at the interface between the output of machine learning models and human interpretation, thereby improving detection and decision quality and increasing reliability and traceability.  
  • The goal of Knowledge Conformity was to develop a plausibility check that would detect decisions made by knowledge-infused AI driving functions that do not conform with formalized knowledge and which can be applied in improving efficiency of AI model training, the reliability of AI deductions and the safety of AI driving functions.    
  • Enabler, Integration, and Demonstration involved evaluating developed functions, components and methods through three selected use cases within a comprehensive system. This holistic approach ensured not only the evaluation of individual components but also assessed their suitability within the broader system architecture.  

The project wrapped up its activities earlier this year and the team presented its results at the KI Wissen Final Event in the House of Logistics and Mobility (HOLM) in Frankfurt on March 21 and 22, 2024. The teams could show tangible results in 17 patents that have been filed, 13 of them by Continental. Looking at the whole project cluster of KI Familie, it was groundbreaking not only in the results it produced, but also in the cultural change of key results sharing across projects. Project partners were able to reap the benefits of the entire project cluster without needing to participate in each individual project. This is highly important and necessary at a time when new players enter the automotive market from the tech company space.

KI Wissen - Corina Apachite - Gilles Mabire - Claudio Longo
KI Wissen Jörg Reichardt
KI Wissen Antje Roya

Collaboration in Artificial Intelligence

Classic automotive questions are resurfacing in the context of AI. Proficiency in artificial intelligence (AI) and its secure integration into contemporary vehicles will define the dominant position in future mobility markets.

The German automotive sector has confronted this challenge through the initiatives of the KI Familie. These projects, characterized by their intricate nature and interconnected framework, foster extensive KI Wissen across various automotive applications.

Learn more about the KI Familie  and their initiatives:

Related Topics

In Berlin, Continental is focusing on artificial intelligence for the mobility of the future.

Continental Officially Opens AI Lab in Berlin; To Support Future Mobility Solutions

Official inauguration of Artificial Intelligence (AI) Lab, which will provide ideal framework and conditions for application-independent development for autonomous driving and robotics beyond our current boundaries

Learn more

Software-defined Vehicle

Our software-defined vehicle approach. Coming generations of vehicles are increasingly characterized by software-enabled functions.

Learn more
car on a digital street with cloud visual and digital bubbles

Do you want to know more?

*If the contact form does not load, please check the advanced cookie settings and activate the functional cookies for the purpose of contact management.