H2 Secure Net

Modern vehicles generate large amounts of sensor data during operation. This can include information such as hydrogen pressure, temperature, energy consumption, fuel or battery levels, and driving dynamics. While this data is essential for condition monitoring and predictive maintenance, it is also subject to strict data protection and cybersecurity requirements. H2SecureNet addresses this challenge by consistently using cryptographic methods. Homomorphic encryption and secure multi-party computation are used in particular to enable calculations to be performed directly on encrypted data, eliminating the need for it to be converted to plain text.

H2SecureNet's system architecture is designed for end-to-end security. Vehicle data is either encrypted at the manufacturer’s premises or directly on the vehicle, before being transmitted to an external analysis entity for processing. There, predictive maintenance algorithms are executed without any access to the unencrypted data being possible. The calculated results, such as condition indicators or maintenance recommendations, are also returned in encrypted form and can only be decrypted by the authorised recipient. This ensures both the confidentiality of the data and the protection of the AI models used.

H2SecureNet uses modern encryption methods to efficiently process large volumes of time-series-based sensor data. The platform supports operations such as aggregation, statistical analysis and basic machine learning processes on encrypted data. Thanks to its modular and scalable architecture, H2SecureNet can be flexibly integrated into existing IT infrastructures and is also suitable for large vehicle fleets. Mechanisms such as end-to-end encryption, secure key management, role-based access controls and audit logs also ensure a high level of cybersecurity and traceability.

Stakeholders

Partners from industry and academia are working closely together to implement the project. The DLR Institute for AI Safety and Security is collaborating with Poppe + Potthoff GmbH and Auto-Intern GmbH. Together, they are developing and evaluating a practical solution for the secure processing of sensitive vehicle data under realistic conditions. The project is funded by the European Union as part of the ERDF competition 'NeueWege.IN.NRW'.

Contribution Institute for AI Safety and Security

The Institute for AI Safety and Security is responsible for designing and developing cryptographic methods and secure system architectures. A key focus is the efficient implementation of homomorphic encryption and multi-party computation for industrial applications. Security mechanisms are also being developed to protect the entire processing chain, including key management, access control, and integrity verification. The ultimate goal is to provide a robust, trustworthy platform that meets the highest data protection and cybersecurity standards.

Project approach

The project partners are taking an iterative approach to development. First, the real-world application requirements of vehicle diagnostics are analysed and suitable data structures are defined based on these requirements. Based on this analysis, cryptographic methods and algorithms for processing encrypted data are then developed and optimised. These are then integrated into a modular platform and evaluated in a test system under real-world conditions with regard to performance, scalability, and security. The results are then fed back into the development process.