Nico Weber

Nico Weber, Patrick Holzer, Tania Jacob et al.

Federated Learning as a decentralized artificial intelligence (AI) solution solves a variety of problems in industrial applications. It enables a continuously self-improving AI, which can be deployed everywhere at the edge. However, bringing AI to production for generating a real business impact is a challenging task. Especially in the case of Federated Learning, expertise and resources from multiple domains are required to realize its full potential. Having this in mind we have developed an innovative Federated Learning framework FACT based on Fed-DART, enabling an easy and scalable deployment, helping the user to fully leverage the potential of their private and decentralized data.

Nico Weber, Christoph Thiem, Ulrich Konigorski

Scenario-based testing is a promising approach to solve the challenge of proving the safe behavior of vehicles equipped with automated driving systems. Since an infinite number of concrete scenarios can theoretically occur in real-world road traffic, the extraction of scenarios relevant in terms of the safety-related behavior of these systems is a key aspect for their successful verification and validation. Therefore, a method for extracting multimodal urban traffic scenarios from naturalistic road traffic data in an unsupervised manner, minimizing the amount of (potentially biased) prior expert knowledge, is proposed. Rather than an (elaborate) rule-based assignment by extracting concrete scenarios into predefined functional scenarios, the presented method deploys an unsupervised machine learning pipeline. The approach allows exploring the unknown nature of the data and their interpretation as test scenarios that experts could not have anticipated. The method is evaluated for naturalistic road traffic data at urban intersections from the inD and the Silicon Valley Intersections datasets. For this purpose, it is analyzed with which clustering approach (K-Means, hierarchical clustering, and DBSCAN) the scenario extraction method performs best (referring to an elaborate rule-based implementation). Subsequently, using hierarchical clustering the results show both a jump in overall accuracy of around 20% when moving from 4 to 5 clusters and a saturation effect starting at 41 clusters with an overall accuracy of 84%. These observations can be a valuable contribution in the context of the trade-off between the number of functional scenarios (i.e., clustering accuracy) and testing effort. Possible reasons for the observed accuracy variations of different clusters, each with a fixed total number of given clusters, are discussed.

Nico Weber, Christoph Thiem, Ulrich Konigorski

While there was great progress regarding the technology and its implementation for vehicles equipped with automated driving systems (ADS), the problem of how to proof their safety as a necessary precondition prior to market launch remains unsolved. One promising solution are scenario-based test approaches; however, there is no commonly accepted way of how to systematically generate and extract the set of relevant scenarios to be tested to sufficiently capture the real-world traffic dynamics, especially for urban operational design domains. Within the scope of this paper, the overall concept of a novel simulation-based toolchain for the development and testing of ADS-equipped vehicles in urban environments is presented. Based on previous work regarding highway environments, the developed novel enhancements aim at empowering the toolchain to be able to deal with the increased complexity due to the more complex road networks with multi-modal interactions of various traffic participants. Based on derived requirements, a thorough explanation of different modules constituting the toolchain is given, showing first results and identified research gaps, respectively. A closer look is taken on two use cases: First, it is investigated whether the toolchain is capable to serve as synthetic data source within the development phase of ADS-equipped vehicles to enrich a scenario database in terms of extent, complexity and impacts of different what-if-scenarios for future mixed traffic. Second, it is analyzed how to combine the individual advantages of real recorded data and an agent-based simulation within a so-called adaptive replay-to-sim approach to support the testing phase of an ADS-equipped vehicle. The developed toolchain contributes to the overarching goal of a commonly accepted methodology for the validation and safety proof of ADS-equipped vehicles, especially in urban environments.

Markus Steimle, Nico Weber, Markus Maurer

To successfully launch automated vehicles into the consumer market, there must be credible proof that the vehicles will operate safely. However, finding a method to validate the vehicles' safe operation is a challenging problem. While scenario-based test approaches seem to be possible solutions, they require execution of a large number of test cases. Several test benches, ranging from actual test vehicles to partly or fully simulated environments, are available to execute these test cases. Each test bench provides different elements, which in turn, have different parameters and parameter ranges. The composition of elements with their specific parameter values at a specific test bench that is used to execute a test case is referred to as a test bench configuration. However, selecting the most suitable test bench configuration is difficult. The selected test bench configuration determines whether the execution of a specific test case provides sufficiently valid test case results with respect to the intended purpose, for example, validating a vehicle's safe operation. The effective and efficient execution of a large number of test cases requires a method for systematically assigning test cases to the most suitable test bench configuration. Based on a proposed method for classifying test bench configurations, we propose and illustrate a method for systematically assigning test cases to test bench configurations in a scenario-based test approach for automated vehicles. This assignment method allows for the effective and efficient execution of a large number of test cases while generating sufficiently valid test case results.

Jonas Ney, Dominik Loroch, Vladimir Rybalkin et al.

Deep Neural Networks (DNNs) are capable of solving complex problems in domains related to embedded systems, such as image and natural language processing. To efficiently implement DNNs on a specific FPGA platform for a given cost criterion, e.g. energy efficiency, an enormous amount of design parameters has to be considered from the topology down to the final hardware implementation. Interdependencies between the different design layers have to be taken into account and explored efficiently, making it hardly possible to find optimized solutions manually. An automatic, holistic design approach can improve the quality of DNN implementations on FPGA significantly. To this end, we present a cross-layer design space exploration methodology. It comprises optimizations starting from a hardware-aware topology search for DNNs down to the final optimized implementation for a given FPGA platform. The methodology is implemented in our Holistic Auto machine Learning for FPGAs (HALF) framework, which combines an evolutionary search algorithm, various optimization steps and a library of parametrizable hardware DNN modules. HALF automates both the exploration process and the implementation of optimized solutions on a target FPGA platform for various applications. We demonstrate the performance of HALF on a medical use case for arrhythmia detection for three different design goals, i.e. low-energy, low-power and high-throughput respectively. Our FPGA implementation outperforms a TensorRT optimized model on an Nvidia Jetson platform in both throughput and energy consumption.