The Munich-based AI startup Deepscenario offers with its AI Scenario Engine A platform for training, testing, and validating autonomous systems of all kinds. The startup focuses on the automotive industry. The platform relies on image processing software that uses traffic scenarios from cameras and thus does not rely exclusively on synthetic data. This solution is intended to enable companies to train, test, and validate their autonomous systems on a large scale. This should enable the startup's customers to bring autonomous products to market in less time and at lower cost. Deepscenario was founded in early 2021 by Holger Banzhaf, Jacques Kaiser, and Nijanthan Berinpanathan.
Co-Founder and CEO Holger Banzhaf explained:
"Deepscenario's technology is industry-leading, and the new funding will enable us to make our solution available to customers worldwide. What sets the AI Scenario Engine apart from other solutions is our scenario mining process. We leverage our groundbreaking image processing algorithms to access representative real-world distributions in all three spatial dimensions and the time dimension."
“Combining the physical world with simulation”
In addition to the High-Tech Gründerfonds and the Mobilityfund, several business angels have also the Munich startup invested. For example Michael Bolle, former CTO and CDO at Robert Bosch and now advisory board member of Deepscenario. He says:
"Training and testing autonomous systems exclusively in the physical world is not feasible, as it requires too much time and resources to achieve the desired results. Simulation is key to overcoming these limitations, but simulators today are based exclusively on synthetic data that do not truly reflect the complexity and diversity of reality. Combining the physical world with simulation is the only way to bring autonomous systems to market with the speed and cost profiles required by industry."
Johannes Weber, Senior Investment Manager at HTGF, added:
"Deepscenario's product is used as a virtual test bed where automakers, suppliers, technology companies, and certification bodies can validate autonomous systems. For example, thousands of driving variations through a challenging two-lane roundabout can be performed, providing important insights into the system's safety and performance."
