Operational Design Domains
Defining the conditions an AI system is designed for—and connecting those limits to data, training, and monitoring.
AI Expert · DLR · Hamburg
I’m Johann, an AI expert and researcher developing safe, resilient AI for aviation at the German Aerospace Center (DLR).
From defining where an AI system can operate to understanding how it behaves—my work connects machine learning with the assurance needed to use it in the real world.
Research Focus
Defining the conditions an AI system is designed for—and connecting those limits to data, training, and monitoring.
Building assurance into AI development, with neural-network analysis, visualization, and certification-oriented engineering.
Bringing learning and control together for collision avoidance, robotics, and human–AI teaming in air traffic control.
Selected Work
A data-driven method reconstructs operational design domains using a deterministic kernel-based representation, with validation through Monte Carlo experiments and an aviation collision-avoidance use case.
Also accepted at TAE and WRL workshops · NeurIPS 2026, Sydney
Johann Maximilian Christensen, Elena Hoemann, Frank Köster, Sven Hallerbach. "Defining Operational Conditions for Safety-Critical AI-Based Systems from Data". Advances in Neural Information Processing Systems (NeurIPS), 2026. Accepted · forthcoming.
A systematic study of Safety Nets explores the trade-off between neural-network size and corrective lookup tables for HCAS and VCAS, reducing storage while preserving correct outputs.
Johann Maximilian Christensen, Thomas Stefani, Elena Hoemann, Frank Köster, Sven Hallerbach. "On the Applicability of Safety Nets: A Safety-By-Design Solution for Certifying Neural Networks". 35th Congress of the International Council of the Aeronautical Sciences (ICAS), 2026.
An aviation AI engineering framework combines EASA’s W-shaped development process with DevOps, supporting iterative improvement and re-evaluation using operational information.
Johann Maximilian Christensen, Thomas Stefani, Akshay Anilkumar Girija, Elena Hoemann, Andrea Vogt, Viktor Werbilo, Umut Durak, Frank Köster, Thomas Krüger, Sven Hallerbach. "Formulating an Engineering Framework for Future AI Certification in Aviation". Aerospace, 2025.
Beyond the Paper
My path runs through aircraft cabin R&D at Airbus, aerial robotics at UC Berkeley, and applied AI at ZAL. At DLR, I bring that experience to safe AI engineering.
I actively acquire research projects through EU and German funding programmes and directly with companies. I lead interdisciplinary projects with budgets of up to €5 million, bringing research partners and industry together to advance safe AI.
Explore My Experience →