21. Internationales Stuttgarter Symposium

MobiLab is following all in all three approaches: the strategy to reach climate neutrality with respect to the mobility of the University of Stuttgart until 2035, the implementation of the car-free campus Stuttgart-Vaihingen and the offer of the Mobility Living Lab Campus Stuttgart-Vaihingen. To reach climate neutrality in mobility until 2035 the University of Stuttgart acts mainly in the areas of the daily mobility of students and employees (commuting) and of the business trips of the employees. To address the students and the employees since the beginning of 2021 the position of a Mobility Manager at the University of Stuttgart is established. The basic idea of the car-free campus Stuttgart-Vaihingen is to shift all parking lots, which are currently scattered on the whole campus to the edge of the campus, finally in one huge car park covering a motorway. Then these peripheral parking facilities, as well as the already existing station of the regional train system (S-Bahn), will be connected with the campus. This will be done with novel means of transport like automated shuttle buses or a rental system with autonomous e-scooters. In this context, forward-looking electrical propulsion concepts in combination with autonomous driving as well as intelligent systems to store and distribute electrical energy are important research areas. Putting all this together, MobiLab finally forms a multifaceted Mobility Living Lab with the three main research topics “research vehicle”, “research street” and “research multi-storey car park”.

Today’s technological change into alternative power trains is having massive impact beyond the vehicle itself. Tests are showing that air streamed electric vehicles are reduced by half in reach towing a trailer, especially a caravan. In order to compensate for the significant loss of traveling distance with a single battery charge EHG, ZF and FKFS are designing a battery electric propelled caravan.In addition to the technical aspects that have already been presented, the approval of powered trailers poses various regulatory challenges. A vehicle category or definition for powered trailers does not yet exist and must therefore be newly created in coordinated association work and in close cooperation with the relevant approval authorities.The additional weight that is created by the drive components represents further requirements in terms of driving license and trailer load. The limit of 3.5 t should be emphasized here. Compensation for the additional weight is therefore necessary.

Future regulations for emissions will increasingly focus on reducing respirable particles, which are caused by abrasion from brakes, tires and roads. PMP (Particle Measurement Program) under the United Nations Economic Commission for Europe (UNECE) has already defined a measurement cycle together with international organizations to measure brake wear on brake dynamometers. Questions about the sampling method and details about the measurement are open.This paper will focus on Brake Emissions as the regulations on the one hand are already on the way, and on the other hand the measuring methods are basically defined. Furthermore the central points are not chemical or legislative details, but a view from system and vehicle perspective.However, the focus of the regulation is currently only on passenger cars with internal combustion engines, since the measurement process only records the emissions of the “brake” component, without taking into account recuperation, for example.The exact definition of the limit values is still pending, but it can be assumed that they will be based on exhaust emissions of internal combustion engines.There are numerous measures to reduce particle emissions, which either aim to generally prevent particle emissions from the brake or to filter the particles that have already been emitted.As a result, the volume of tests relating to the brake will increase massively, as all variants have to be tested and certified. On the other hand, the necessary changes to the brakes are also affecting other properties of the entire vehicle, causing e.g. compromises between performance, emissions and comfort.The automotive industry will then face a further increase in the challenges when considering tire and road abrasion emissions.

Stricter emission and consumption requirements lead to an increasing importance of hydrocarbon (HC) and carbon monoxide (CO) emissions for diesel engines. On the one hand, the amount of HC and CO raw emissions are important for emission prediction as well as for exhaust gas after treatment. On the other hand, they are also important for fuel consumption prediction, since part of the fuel energy is bound in these molecules. Because of this importance, a phenomenological HC/CO model was developed to predict raw emissions for diesel engines. The HC/CO model takes the main emission sources into account: Fuel that slowly leaks out of the injector after the end of injection is included in the HC model. Furthermore, HC and CO emissions from cold areas near the combustion chamber walls are considered. HC and CO emissions from lean overmixed spray regions are also implemented. As an additional source, the CO model considers locally understoichiometric regions of combustion. The validation results of the model using two passenger car diesel engines are shown for this purpose. With the help of the calibrated model, the shares of the different emission sources in the total HC and CO emissions of an engine can be estimated and consequently optimization potentials can be identified. An application of the CO model in the context of an acceleration process within a real driving emissions (RDE) simulation is shown. In this context, it is demonstrated how transient influences on total emissions and emission sources are reproduced in the model.

Permanent magnet excited (PMSM), externally excited synchronous machines and asynchronous machines (ASM) are state of the art in vehicle traction applications. Currently, the externally excited synchronous machine (EESM) has a high market share in Europe and also asynchronous motors are encountered in many battery electric (BEV) vehicle models still. However, the PMSM becomes the dominant type of vehicle traction drive in the product portfolios of the leading car manufacturers. The reason for this trend is that PMSM machines deliver torque at high efficiency over a wide range of load situations with no need for expensive rotor cooling in the typical load profile of a road vehicle. The main drawback of PMSMs: the mining of rare earth materials produces very harmful greenhouse gases. In the recent years, attempts were made to advance the switched reluctance motor (SRM) in order to make it suitable for traction applications. The advantage of SRM is that no permanent magnets are needed and its production cost is low. The most significant disadvantages are high torque ripple, noise, low efficiency in partial load situations and necessity for expensive rotor cooling. In this paper, the approach of a novel motor concept is shown that relies on seven phases. It overcomes the common drawbacks of SRM is a very promising concept for the next generation of electric vehicles.

During the dimensioning process of hybrid electric propulsion systems, engineers are confronted with a large variety of possible components. A key influence on overall system efficiency is the electric drive system (EDS). Usually, the comparison of different EDS is limited to comparing loss maps and visualizing their differences. Since loss maps are not easy to handle, a simple model of the efficiency characteristics can be helpful to identify the best component for the specific application. This paper proposes a simplified physics-based approach to model the loss characteristics of an EDS based on polynomial functions. The EDS consists of a permanent magnet synchronous machine (PMSM) and an inverter (INV). The developed model achieves good accuracy for both, the base-speed region and the field-weakening region. The modeled losses are dependent on DC-bus voltage in motor as well as in generator mode. The proposed method allows the description of the EDS efficiency characteristic by only few scalar coefficients and increases thereby the comparability of different EDS.

The implementation of artificial intelligence (AI) systems in automotive software development still is an obstacle. Despite of accelerating scientific research and big wins in this field, the practical application is only possible in restricted environments or non safety critical components. There is a need to develop methods to verify the robustness and safety of AI software modules. The data based generation of deep learning (DL) algorithms creates black box models, which properties inhibit a validation as it is done for deterministic algorithms following ISO 26262. This paper introduces methods to assess the plausibility of AI model outputs. A description of the training data domains for a robust training is accomplished by means of one-class support vector machines (OCSVMs). This anomaly detection process encloses valid data within a DB, to be able to verify model outputs during operation. A further categorization of the training data domain into 20, equally spaced sub-domains led to best results in detecting implausible model calculations.

Reducing the energy consumption required for heating ventilation and air conditioning for passenger compartment is one of the key challenge for the thermal management system of electric vehicles to extend the driving range. Particularly at low ambient temperature the power demand for cabin heating increases and drains the battery capacity at the cost of the driving range. Due to the high air exchange rate of the vehicle cabin to the environment, a large share of the heated air is released unused into the environment via the rear exhaust vent as warm exhaust air. Consequently, the heat contained in the cabin air remains unused. A new approach considered in this study is to use the thermal energy in the cabin air as a heat source for a heat pump system to reduce the heating power required to heat the fresh intake air from the ambient. In vehicles with rear air-conditioning systems, the basic components, evaporator and fan, are already present. In this study, the potential of waste heat recovery from cabin exhaust air is demonstrated with the help of simulation and evaluated for various boundary conditions. Limiting effects on the performance are identified and critically discussed.

Providing clean air for PEM fuel cells is crucial to prevent premature degradation of the stack caused by harmful gases poisoning platinum catalysts. For this purpose, special activated carbon mixtures are used that reliably adsorb target gases such as NOx, NH3 and SO2. Basic investigations in research projects create the basis for the application-oriented design of the filter elements. To prevent flooding of components in the cathode air path including the stack, water separators are used at various positions in the overall system. The components in the air path are designed according to the temperature and pressure requirements prevailing in the application, with sensors and actuators enabling system control and management. Applying optimized silencers, disturbing noises, e.g. from the compressor, can be reduced or even eliminated. Central system components such as humidifiers and heat exchangers positioned in charge air lines can be efficiently integrated into the overall system. Additional water separators to protect the turbine and prevent the discharge of liquid water from the tailpipe in the cathode exhaust path, as well as a reservoir, are advantageously integrated into the overall system using plastics technology. This holistic approach enables acoustically optimized and compact cathode air paths for PEM fuel cells.

Fuel cells will play an important role in reducing CO2 emissions from transport, especially for heavy duty use cases. Current fuel cell systems have an electrical efficiency of ~50%. The remaining 50% of energy lead to heat generation, which have to be removed by the coolant loop at very high flow rate. In order to ensure a robust running system and to avoid electric shorts, the electrical conductivity in the fuel cell stack has to be kept low. By using the ion exchange technology, ions, which are released by several components in the coolant loop, are separated. In addition to highly-efficient ion removal, low pressure drop, mechanical robustness and costs are the challenges in the development of such ion exchangers. MANN + HUMMEL has developed a customer independent design, which is available as market product and covering most of current system requirements. This paper summarizes the product design and testing on ion exchange resin and product basis as well as the MANN + HUMMEL market product approach to reduce systems costs by standardization of components. Furthermore, a brief outline of a particle filter for a FC coolant loop is given.

A great challenge in advanced engineering of mechatronic products in automotive applications is an efficient way to solve a conflict of objectives: On the one hand a high degree of flexibility is necessary due to frequent changes of requirements. On the other hand, keeping the traceability is required to guarantee Automotive SPICE (ASPICE) conformity. A particular challenge is the use of a heterogeneous verification tool landscape from various specialized suppliers. Most of the test runs require setups with flexibly scalable measurement devices from additional suppliers.MAHLE encounters this challenge by deriving project specific test cases out of a global requirements database which are used consistently at each system integration level of a mechatronic product. Simulations for Model-in-the-Loop (MiL) and Software-in-the-Loop (SiL) as well as Hardware-in-the-Loop (HiL) and dyno testing use the same test procedures out of a superordinate automation database which does not differentiate between the levels of system integration and maturity. This leads to standardized test procedures at all system-specific, model-based verification steps and allows a comparability of the results after standardized post-processing during advanced engineering. In this paper, for the first time a comparison of the results between multiple integration levels with heterogeneous tools is presented on the example of an electric traction drive system. Thus, quick V-model iteration loops with instant findings about the mechatronic product in development are enabled.

Due to limited space and the focus on cheap and effective power electronics in automotive vehicles, the demand for fast switching devices has increased drastically. Such fast switching devices produce high electromagnetic interference (EMI) emissions which result in a gain of size and weight of EMI filters to comply with the electromagnetic compatibility (EMC) requirements. Due to uncertainties, the EMI filters used are often oversized. The goal of this paper is to show how an EMI filter is designed that fulfills the requirements in a way that smaller components can be chosen. Therefore, a passive EMI filter for a 400 V to 12 V DC/DC converter is created. A common method used to calculate the L and C filter components is to find the minimum required corner frequency and determine appropriate parts. However, this is a simplification of the filter design. In reality the source and load impedances also influence the design. Those impedances are often defined as 50 Ω. But the real source and load impedances of the system deviate from 50 Ω. In order to achieve effective filtering of the emissions, the knowledge of the noise source impedance is a very important factor in calculating the exact insertion loss of the filter. In this paper, the value of the impedance is estimated by analyzing the emission paths of the DC/DC converter. Then the influence of the impedance on the filter design is shown. With that knowledge, an improved EMI filter is created and demonstrated with measurements.

In order to let even big and heavy SUV vehicles fulfill the emission targets of the future a PHEV-hybrid- drivetrain for propulsion is absolutely necessary. The restrictions of the new Euro-7-emission limits and the special WLTP-testing and -calculation regulations enforces special optimizing strategies of the hybrid drive which will be discussed intensively. On basis of WLTC- and RDE-measurements the presentation shows the optimum application of a hybrid drive which ensures an emission behavior within the legislation corridor.

From the view of the energy provider integration of e-mobility into smart grid is important, which includes effective dynamic load management and the use of high voltage-batteries in the vehicle to store and feedback energy. The user expects sufficient and reliable charging points to recharge his electric vehicle everywhere and at any time. Professional solutions depend on local conditions and actual request for charging power and energy. For expansion of power grids several studies and recommendations exist already. If controlled charging is implemented expansion of the local distribution systems will be moderate. A communication system connects all partners in the system. It supports all functions for proper grid integration of electric vehicles and charging solutions: charging control, load management, authorization, billing system, bi-directional charging, e-roaming and value added services. Working together with enterprises in various sectors, the automotive industry is in a position to make a significant contribution to our transition to a clean-energy economy, with systems that also propel digitization. In doing so, we can effectively support our climate objectives at a national and international level. These system applications also offer great potential in the new business sectors of digitization.

An ongoing market penetration of battery electric vehicles will demand charging infrastructure at places without access to energy grid (e.g. car parks at events, touristic infrastructure, campsites, parking lots at hiking spots). This demand exists for short time at events and seasonal at campsites. State of the art power supply are diesel powered energy sets to build up short term charging infrastructure. The DLR, together with Vector Informatik GmbH, develops a mobile charging station for electric vehicles, which can be refilled with gaseous hydrogen in order to provide local CO2 free charging infrastructure without grid connection.This paper will present the concept of this charging station for applications in Baden-Wuerttemberg. Its features will be minimum two Type 2 charging plugs for electric vehicles and minimum two charging plugs for pedelecs, which will be integrated into a car trailer with maximum 750 kgs. The authors focus on the operating principle for the mobile charging station, the design and integration into the car trailer and the intelligent charging system. Finally, further applications will be derived from the “proof of concept” in the outlook.

With the appearance of the autonomy levels (according to SAE J3016), the responsibility of the driving ability is iteratively transferred to machine systems. An edge case is countered by a human driver with improvisation, while machines try to use known situations or given conditions to induce a human equivalent behavior. Such a system is based on landmarks like traffic signs, markings or structural conditions – the conventional road infrastructure. If there are not enough features to assign a situation, the system must fail. This infrastructural dependency, which future vehicles with increasing levels of autonomy must be able to cope with, was examined using the SLAM (Simultaneous Localization and Mapping) problem as an example. Two problems exist here: the landmark association and the diverging sensory inaccuracy. This contribution solves the association by proposing machine-readable traffic signs. The change of the conventional infrastructure allows the introduction of a conceptual fallback matrix, which in this context may lead to a decoupling of infrastructural dependency. This leads to the thesis that the vehicle of the future will no longer be defined by its engine power, but rather by its sensor technology and computing power to provide its autonomy.

Artificial Intelligence offers great potential for automotive development because it offers the opportunity to develop new or enhanced functionalities. To make that a reality, automotive development processes and methods need to be enhanced. In this paper we show how inductive learning instead of programming of functions changes development and how the requirements for this need be expanded. Furthermore, there is a need for a specific AI test strategy and test execution. We will also argue the need to further reduce the duration of development cycles and to increase the flexibility of resource allocation. In addition, the provision of data becomes a basic building block of successful AI development, which needs to be planned and managed properly. All these changes can become severe challenges. Best practices and experiences will be used in a reference model for the quick and detailed assessment of AI based automotive development. This way, the necessary changes can be implemented and mastered much faster and the benefits from AI through can be utilized much earlier.

In the automotive industry, high expectations have been placed on the industrial application of Artificial Intelligence (AI) for several years. These expectations focus on both the development of vehicles and the vehicles themselves. Machine Learning should help to accelerate processes, increase quality, and even pave way to tackle previously non-feasible challenges. Indeed, AI methods hold the potential to meet these needs in most cases, provided they are applied in the right way and used in the correct context.This paper gives a descriptive overview of AI methods and their practical application in industry and research projects, especially in vehicle development. TWT is constantly developing new successful AI-based solutions in customer projects. In addition to various abstracted examples from industrial projects, exemplary results from TWT's participation in the research project AutoAkzept are presented.

In recent years, more and more automobile manufacturers and suppliers have started to focus on the development and practical application of objective driveability evaluation systems. The limited reproducibility and comparability of subjective driveability evaluation is significantly improved by using objective driveability evaluation systems. A widely used system for driveability objectification is AVL-DRIVE. This system calculates a real-time rating, so called AVL-DRIVE rating (DR) for comfort and dynamics in all relevant driving conditions of the vehicle. In addition, physical parameters are calculated, so that developers can quickly gain a deeper understanding of the driveability evaluation and optimize driveability by analyzing the physical parameters and ratings.For further increase of efficiency in driveability calibration, additional relevant physical parameters and objective evaluations based on xCU signals are calculated in AVL-DRIVE, subsequently, tool-based correlation and optimization studies are carried out with the support of AI (Artificial Intelligence). The aim is to support the calibration engineer in the data analysis to quickly identify abnormal events in the calibration and determine the most likely root cause. This new solution was developed and implemented in cooperation with the Institute of Automotive Engineering (IAE) at the Technical University of Braunschweig and AVL List GmbH.Particular attention is paid to the linking of poor driveability rating (DR) and underlying settings in the calibration. In particular, the torque intervention, the clutch pressure control, and the speed adjustment during the shift process in automatic and dual clutch transmissions are considered. With the developed AI-based methodology, the efficiency of objectified driveability calibration of vehicles can be increased.

Due to the electric offensive and the resulting new freedom in vehicle concept development and design, it is necessary to leave the evolutionary development track and instead take a revolutionary approach in order to inspire customers with innovative interior and powertrain concepts. To accelerate the revolutionary development and to be able to guarantee more flexibility, it is necessary to make decisions early in the pre-development of vehicle concepts. Up until now, powertrain concepts, interior concepts and exterior designs have been modelled according to requirements, and then thereafter, combined to form a vehicle concept. The coordination between the areas may only take place relatively late in the development process. This may lead to an increased need for coordination and correction, more work, and possibly longer development processes. In order to achieve higher efficiency in the development process, a vehicle topology concept method has been developed. This work deals with the packaging of given powertrain- and interior-concepts and the exterior design, as well as the validation of this overall package. The method determines the compromise between modules, which results in the least amount of restrictions of the overall vehicle concept. It is based on an abstract vehicle model, which arises from use cases and dimensional chains. The resulting diversity of possible concepts aids the decision-makers in strategic decisions as to which concepts will be further developed or transferred to series development.

Mobility requirements are changing, and with them the demands on mobility carriers and infrastructures in terms of flexibility, reliability, cost-effectiveness, energy efficiency and environmental compatibility. Because of increasing digitization in recent years, new solution approaches and business models are coming into question, for example, through new types of automated and networked vehicle systems and functions. The IT advances enable a complete rethink of the entire mobility system with all associated subsystems such as the vehicle.The opportunities that digitization offers for the mobility system and the conceivable future scenarios for the mobility of tomorrow are currently being investigated in the “Zukunftslabor Mobilität” research project. This paper briefly reports on the project and systematically examines the influence of digitization on the mobility system. In detail, the various factors influencing the mobility system are discussed. The current situation and the alternative development paths or projections for the mobility of tomorrow are systematically recorded, taking into account the digitization trends. Subsequently, developed future images and applications are derived, as well as a short outlook on future activities is given.

The automotive industry is experiencing fundamental changes due to the current developments in electrification, connectivity and automation. Companies are therefore faced with the question of which technologies and products will be relevant in the future. This applies equally to OEMs and the entire supplier chain. With the technology calendar TKBW presented here, the expected technological developments of passenger cars are described until 2035 on the basis of more than 40 modules, i.e. technology groups, with almost 150 technologies in the areas of propulsion, connectivity and automation. It is thus significantly more comprehensive than other roadmaps and is aimed at small and medium-sized companies that need to strategically align themselves in the technology shift. Maturity of the technologies is characterized by technology and manufacturing readiness levels. This paper describes the methodology used (including roadmapping, Delphi survey and patent research) and explains by means of exemplary findings how the results can be used to find new technology and business opportunities. The work resulted in three products, which can be accessed at www.tkbw.de : a guideline describing the big picture, a module catalog presenting technology roadmaps for each module, and technology fact sheets as a structured reference book.

By separating the drive unit (driveboard) and the transport capsule, the driverless and electrically driven vehicle concept U-Shift enables a new kind of flexibility and efficiency for the mobility of tomorrow. In combination with various types of capsules, the driveboard is used to transport both people or goods. The capsules can be changed quickly, drives quietly, without the need for operators and without being tied to changing stations. Special capsules enable a wide variety of business models and applications - from barrier-free on-demand buses to parcel delivery services. Cost and resource efficiency are key development goals, which is why expensive systems such as the complex technology for driverless driving are primarily housed in the driveboard. In a first step, a rollable prototype was built. Development of a further prototype with extended functions for driverless driving has already begun. We present our vehicle concept U-Shift with the special features to the different vehicle technologies. The partners of the DLR Institute of Vehicle Concepts in the U-Shift project include the Research Institute of Automotive Engineering and Vehicle Engines Stuttgart (FKFS), which is developing the powertrain, the Institute of Vehicle Systems Technology (FAST) and the Institute of Information Processing Technology (ITIV) at the Karlsruhe Institute of Technology (KIT), which are contributing the chassis and the electrical/electronic architecture, respectively. The Institute for Measurement, Control and Microtechnology (MRM) at the University of Ulm is responsible for the automation.

Based on trajectory planning algorithms, a new, real time capable control algorithm for active suspension systems is introduced in this paper to face new requirements on vehicle body motion in low frequency range. The algorithm makes use of pre-view information about the road profile and the future trajectory of the vehicle to find an optimal sequence of control signals for a given path. Optimality is determined via a nonlinear cost function. In the first step, requirements on low frequency motion for autonomous vehicles to are derived from the expected usage scenarios, including side activities like working. Therefore, existing functions like curve tilting are reviewed and requirements are compared to comfort criteria for long-distance trains. This is followed by a detailed description of the proposed controller and two implementations for online and offline optimization. The performance of the controller is then evaluated using a simulation environment with a 10 DoF vehicle model, explaining the influence of different cost functions and investigating model accuracy. A simulation of different control algorithms including state of the art non-preview and preview controllers shows the advantages of the new method.

Autonomous vehicles are expected to push towards the evolution of mobility environment in the near future, by decreasing traffic accidents, traffic congestion and vehicle fuel consumption. One of the main limitations they face is motion sickness (MS) and it seems capable of putting in risk fully automated vehicles wide impact, as well as their acceptance by the public [1, 2].

This paper outlines a research project at FKFS aiming for systematic objectification of the passengers comfort in highly automated driving situations. Therefore, a modular vehicle controller model as well as realistic and comfort-relevant driving scenarios were implemented to complement the virtual test environment of the Stuttgart Driving Simulator. The real time capable controller allows simple online adjustments of its control behavior. Besides the controller, an exchangeable, commercially available vehicle model represents the plant vehicle in the simulators real time simulation network. The environmental interaction and traffic vehicles are also implemented with the help of commercial or proprietary tools. The relevant test-scenarios are iteratively determined and created. They contain mainly motorways, country roads and appropriate connections in accordance with German road construction guidelines. The 3D-Road-Surfaces consist of measurement data, synthetic random surfaces and hybrid variants. The measurement data used maps certain motorway parts, which are particularly relevant for comfort. As the measurement data contains only two tracks, an extrapolation method expands it to a 3D surface by using synthetic shapes in lateral direction. This approach turned out to keep the relevant road specific singularities, like transverse grooves in the roadway, while representing a realistic surface for a whole motorway with several lanes. First project results indicate that not only the control-systems behavior or the travelled scenario affect the passengers comfort, but also interdependencies between controller and scenario have a relevant impact. Pursuing the projects target of a comfort assessment method for highly automated driving functions, a subject study is currently in progress to gain subjective comfort ratings with statistic relevance. Due to the found indications, the study will focus on the lateral trajectory planning for motorway lane changes. More precisely, the passengers comfort on asymmetrically shaped lane change trajectories will be evaluated in contrast to symmetrical shapes and in dependence on different driving situations. As a next step towards the aimed comfort assessment method, the captured subjective ratings are to be correlated with objective criteria further on.

For autonomous cars it is crucial to perceive its current environment to ensure safe driving maneuvers. Light detection and ranging sensors (LiDAR) are often used for object detection due to their accurate distance measurements. However, point clouds sensed by LiDAR provide information of the environment which are not important for object detection algorithms (e.g.: vegetation, buildings). Adding semantic segmentation information to the point cloud supports object detection algorithms and improves their performance.Within this work we transfer well established semantic segmentation methods from the image domain to point clouds and evaluate the performance on solid state LiDAR data. We successfully show the applicability of semantic segmentation methods on this new sensor technology. Furthermore, we compare semantic segmentation approaches which operate on different input representations and discuss the benefit of additional information like intensity measurements on the algorithm’s performance. The evaluations are conducted on solid state LiDAR measurements from German highway scenarios.

Compared to standard navigation maps, HD maps contain precise additional information for automated vehicles. To exploit this information, a lane-level accurate localization estimate within the HD map needs to be available. Here, the computation and memory overhead of the localization algorithm needs to be as small as possible to enable a reasonable usage of HD map information. Since the accuracy of GNSS measurements is insufficient for an accurate localization at lane-level, HD maps commonly contain geo-referenced landmarks like lane markings and boundaries that can be detected by series-production camera sensors, especially on highways. These measurements are fused with the motion measurements provided by the ABS and ESP sensors of the vehicle within a Bayesian filtering framework. Here, the most popular localization algorithm coming from the field of robotics is the particle filter (PF). However, for a robust estimation, a sufficiently high number of particles is required which is in contradiction to the objective to minimize the computational and memory overhead of the localization algorithm. Hence, in this contribution a memory and run-time efficient Extended Kalman-Filter (EKF) based solution is proposed employing measurements from series-production vehicle sensors and a public HD map from two highways near Düsseldorf. Experimental results achieved with the EKF are comparable with the localization accuracy provided by the PF while run-time and memory consumption are considerably reduced.

With the growing automation of road vehicles, the demands on their systems for environment perception increase. As one component the knowledge of one’s own pose (position and orientation) is of great importance. In the project Absolut a highly automated road based People Mover is being established and operated on a route in Leipzig. Among its focusses is the development of a localisation system that provides robust pose estimation under variable ambient conditions. This includes locations with GNSS shadowing and high environmental dynamics. The resulting system relies on information from GNSS, vehicle- and LiDAR odometry. For areas of the route where GNSS positioning cannot be guaranteed a georeferenced point cloud map is generated with offline Simultaneous Localisation and Mapping (SLAM). The paper focuses on a method to locate the vehicle in this map using a $$360^\circ $$ 360 ∘ LiDAR. The approach is able to determine the global pose of the vehicle online to compensate GNSS-outages. The registration of the current scan in the map is done using the Normal Distribution Transform (NDT). A superimposed particle swarm optimisation (PSO) ensures a fast convergence of the initial state. The summary contains first experiments on the systems performance, as well as an outlook on key points for further development.

The advancing automation within the mobility sector poses new challenges. The open parameter space of potential traffic scenarios turns out to be difficult in the development and certification of advanced driver assistance systems. Scenario based, simulative validation of driving functions appears to be a promising solution. Given the assumption that only a fraction of all traffic scenarios is safety critical and should be considered for the evaluation of driver assistance systems, a simulation based selection of test relevant driving scenarios can be carried out. With realistic sensor models available the virtual testing of driver assistance systems is cheaper and faster than conventional test drives. Phenomenological sensor models do not require detailed environment models and therefore compromise accuracy and effort. The objective of this work is the development of a phenomenological LiDAR sensor model that reproduces the actual, measured detection capability of LiDAR sensors. Avoiding empirical radar backscatter cross sections, that strongly distort the detection capability of conventional LiDAR sensor models and mapping the measured detection capability onto the phenomenological LiDAR sensor model promises enhanced model accuracy over traditional phenomenological modeling approaches.

In the course of the rapidly changing automotive markets, the demand for innovative and high-performance vehicle system solutions is increasing considerably. In this context, a cross-domain development methodology focused on customer needs (UX) via so-called features (vehicle functions that can be experienced) is also becoming increasingly important.The feature-based development approach is a cross-domain development methodology that is consistently oriented towards customer needs (UX). In order to master the complexity of a complete vehicle development including all interactions between the individual subsystems, the feature-based development approach focuses strongly on methods and tools from model-based system engineering (MBSE). This enables the holistic development of individual vehicle features along their functional and physical chains of effect and takes into account all interfaces and interactions on individual subsystems as well as components. This makes it possible to derive concrete requirements for the subsystems and components of an entire vehicle based on “high-level vehicle use cases” and to generate functional component specifications from them. In the further course of vehicle development, individual components and subsystems up to complete module groups and rolling chassis are developed in such a way that they represent an integrated and coordinated overall solution in the vehicle network.With feature-based development, Bosch Engineering is able to offer high-performance vehicle system solutions that are consistently aligned with customer requirements.

As cars become smart and connected, automotive development gets more complex. Traditional disciplines such as mechanical, electrical and software engineering can no longer remain separate, but are required to collaborate. Further, new regulations and standards require engineering activities across domains to be traceable and their outputs to be consistent. Current IT-architectures do not sufficiently support those new needs. Automotive engineering domains mainly rely on separate expert tools. Interfaces are, if at all existent, inefficient and come along with a loss of data and information. Sophisticated integration of IT-systems is the basis for solving the drawn problems, as it enables efficient and effective collaboration amongst domains. Hence, it also forms the basis for trending approaches in automotive engineering such as model-based systems engineering and function-orientation. Thus, a state-of-the-art integration architecture for the automotive engineering IT is presented in this paper. For instance, it suggests replacing former implicit relations between assets of separate systems by explicit ones through applying an ontology-driven Linked Data approach. This improves traceability and enables referencing, addressing and reusing IT-assets across expert tools. Therefore, it is further proposed to exchange data efficiently among systems in an innovative API-driven manner. The holistic integration architecture enables effective and consistent collaboration and ensures traceable work results across domains. The result is a future-proof architecture that serves as a blueprint for automotive companies to face the challenges of developing innovative cars as of today and tomorrow.

With the necessity to a further reduction in fuel consumption or CO2, the late combustion phase is of high importance throughout the range of operation. Because of the high differences in density between burnt and unburnt zones, a relatively large unburnt mass is concentrated in a relatively small volume near the wall and the top land volume.To investigate the thermodynamic interactions between the boundary zones of the combustion chamber and the flame front, a state-of-the-art Mercedes-Benz-single-cylinder research engine was equipped with extensive measurement technology. To detect the piston movement and the top land volume, eddy-current distance sensors were applied to the cylinder. The radial pressure distribution in the top land volume is detected with four miniature pressure transducers. As main part, surface thermocouples were applied in two layers around the top land volume with additional surface thermocouples in the boundary zones of the cylinder head as a reference.Based on a wide range of operating points and a spark timing variation, the temperature distribution in the different top land levels is investigated in comparison with the cylinder head temperatures. Lastly, two different top land volumes, realized by a piston variation, are compared. Thereby, the temperatures in the bigger top land volume indicate a slightly lower compared to the standard geometry. This difference even increases for the lower top land level.

The intended aggravation of the emission legislation for commercial vehicles in the USA and in the EU as well as unavoidable compromise between fuel consumption, NOx and soot emissions require completely new approaches to find a multi-criterial optimum.IAV developed a three-dimensional shaped piston bowl for the installation in a single cylinder heavy-duty engine. The purpose was to improve the mixture formation by increasing the spray penetration length and to avoid spray-wall interaction as much as possible. Coincidental the surface temperature of the piston crown was increased to reduce the piston-side wall heat losses and consequently increase the thermal efficiency of the engine.The publication comprises the design dimensioning and the additive manufacturing process of the piston with which the piston crown and its thermal insulation can be achieved in a beneficial way. Furthermore, the results of the mechanical testing and some thermodynamic differences to the referenced series piston design are presented.

In recent years, the international emission standards for vehicles have become increasingly stringent, introducing new homologation cycles and limiting emission worldwide. In the next future, CO2 target and RDE regulation will force engine manufactures to further increase the efficiency of the engines making them operate with air to fuel ratio ( $$\lambda $$ λ ) equal to 1 on entire calibration map. For gasoline engines, the knock tendency is the main barrier for the operation with stoichiometric mixture at high loads, which is an important driver to decrease fuel consumption and improve catalytic converter efficiency. One possible way to prevent knocking is to reduce the engine combustion chamber temperature. This paper presents an optimized hollow head valve able to improve knocking tendency by reducing significantly valve combustion face temperature. An innovative design and manufacturing process enabled to achieve outstanding cooling effect, as demonstrated by thermometric testing. Engine calibration testing with hollow head valves proved a remarkable gain in $$\lambda $$ λ and spark advance, if compared to hollow stem. This effect translates into a clear advantage for engine manufacturers, enabling them to improve fuel consumption.

The necessity to improve the Noise Vibration Harshness (NVH) performance and, in particular, to reduce the flow noise produced by turbulent exhaust gases is requiring automotive manufacturers to optimize Internal Combustion Engines (ICE) exhaust systems from the early design phases. Computational Aeroacoustics Analyses (CAA) can be seen as a promising tool to support engineering design and optimization, enabling the evaluation of system performance and allowing the selection of effective noise reduction strategies.This work deals with the application of a CAA methodology for the assessment of noise mitigation potential of two heavy-duty exhaust system designs. Firstly, the mean flow field is obtained with a steady RANS simulation performed with the 3D-CFD code Star-CCM + . Then, the acoustic software ACTRAN SNGR is used to synthetize the noise sources and compute the radiated acoustic field. Flow noise level predicted with the hybrid approach has been compared with the results of a Direct Noise Calculation (DNC) performed in 3D-CFD for the baseline design, to determine absolute levels. It follows a fast optimization loop with the hybrid approach to evaluate the noise attenuation obtained thanks to a design modification in the exhaust diffusor. The proposed methodology significantly reduces the timeframe required to assess exhaust system’s performance, reducing the need of 3D-CFD DNC simulations, and is compatible with the development of new products in a commercial context.

SI engines are typically designed to operate at the knock boundary, as engine knock limits their efficiency and thus further reduction of CO2 emissions. To consider this phenomenon in an engine process simulation, a precise knock model is needed. First, since auto-ignitions precede knocking events, the introduced 0D/1D knock model predicts when the unburnt mixture auto-ignites. For this, the model considers the two-stage auto-ignition of gasoline fuels, which occurs at specific boundary conditions, by calculating two Livengood-Wu-integrals. The model’s auto-ignition prediction is validated against measurement data of a single cylinder engine covering almost 3 000 knocking single working cycles at broadly varied operating conditions.To determine the operating conditions where the knock boundary is reached, the newly developed knock criterion considers the state of pre-reactions of the unburnt mixture before the start of combustion. For its validation, simulations were performed to compare the point where 50% of the mass is burned (MFB50) of each simulated operating point to the MFB50 of the corresponding measured average working cycle at the knock boundary. Though the introduced criterion is calibrated at only two operating points, this new knock model predicts the MFB50 at the knock boundary at broad variations of operating conditions very precisely with a mean deviation of 1.55°CA, helping to improve the development of SI engines in 0D/1D simulation. Furthermore, a three-parameter-approach to model the knock frequency is introduced.

Due to the increasing pressure on vehicle emission reduction, the development of powertrains is changing considerably. Hybridization offers a significant potential in this context, leading to increasing powertrain complexity, and thus to an increase in the number of the control functions to be developed and tested. Therefore, a seamless validation of the concepts in early development stages is necessary to enable cost and development time optimization. For these reasons, the virtualization of powertrain components has become a necessity in the development process, and simulation methods for concept and testing gained fundamental importance. In this context, this paper proves the potential of X-in-the-Loop (XiL)-based methods. In particular, the role of Model-in-the-Loop (MiL) methods in early stages of the development process is highlighted as a cost-effective way for the definition of the most efficient hardware and software functions and for the development of new hybrid powertrain architectures. As one example, a methodology based on XiL simulation and its heterogeneous model landscape of different powertrain topologies will be presented. Specifically, a MiL setup and the corresponding Engine-in-the-Loop (EiL) platform of a parallel hybrid (P2) powertrain have been successfully used for the development and testing of a new emission-optimized control strategy. The results show improvements of the strategy robustness, and significant benefits in terms of emissions reduction.

An unobscured view from the vehicle during rainy weather conditions is essential for occupants’ safety and comfort. With the decrease in the time available for vehicle development and testing, it is becoming even more important to control and predict vehicle water management early in the development cycle to avoid undesired soiling effects. To do so, a transient external aerodynamics airflow coupled with a discrete particle phase is simu-lated, where rain droplets away from the vehicle are modeled as Lagrangian particles and rain droplets that impact the vehicle are represented through a film model. A novel, time-efficient implementation of a windscreen wiper based on [1] within iconCFD [2] is also summarized. A representative example - rain-induced reduced visibility on the side window - is considered to discuss the physics, time scales, and recommended best practices for rain soiling simulations. The simulation results for this example case are compared with experimental data.

To prevent rollover is one main goal in the development of passenger cars. The early consideration and simulation of those vehicle properties show high potential to support the serial development. The advantages lie in the reduction of efforts through systematical deduction of countermeasures and thus a reduction of risks due to changes in late phases of the development. Further, the use of systems engineering methods supports the traceable and structured deduction of technical solutions. The presented paper shows a generic process to develop technical solutions in order to reach the defined vehicle rollover properties using systems-engineering and a simulation based approach. In this context a simulation environment is built up and validated. The process is exemplified with the development of a specific vehicle project.

In this paper, an analysis of a subjective evaluation of manually and automatically executed longitudinal and lateral parking maneuvers using Subjective Assessment Indicators is presented. With the introduction of autonomous driving, parking maneuver assistants are essential functional components. Driver assistance systems will only be accepted if they perform decisively better than the human driver. Whether the performance of such a system meets expectations is ultimately a subjective impression. For this reason, an analysis of the parking performance of humans and parking assistant systems is carried out based on a new innovative subjective evaluation method. This new subjective evaluation method is based on the so-called Subjective Assessment Indicators which cover the relevant areas of a parking maneuver but still do not reach a level of detail that makes evaluation unsuitable for customer. Using the new subjective evaluation method, a driving study was conducted with 21 participants and two different test vehicles. The participants evaluated both manual and fully automated longitudinal and lateral parking maneuvers purely digitally using an evaluation app. As the results of the study show, parking assistants still have notable deficits compared to human performance in some evaluation areas and show considerable potential for improvement. As the subjective evaluation method used is suitable for all parking maneuvers and vehicle types, the results of this and potentially further studies form the basis for determining Key Performance Indicators for parking maneuvers. This enables virtual development of automated parking systems, as a link can be established to subjective customer evaluations.

Increasingly stringent exhaust emission regulations and CO2 fleet emission targets are leading to a continuous electrification of vehicles. The market share of battery electric vehicles (BEVs) is increasing significantly due to the growing supply, sales promotion programs and environmental bonuses, as well as increasing awareness of environmental issues among end consumers. The key criteria for end consumers when it comes to the procurement of a BEV are range, charging time, charging infrastructure and vehicle costs; here, vehicles with conventional drives have an advantage.Although electric powertrain configurations and their higher “tank to wheel” efficiencies allow efficiency improvements, increases in electric ranges mostly require larger and thus heavier batteries. Although this usually leads to low centers of gravity, additional functional measures are needed to make the higher vehicle weight less noticeable.The focus of this paper is on introducing a methodology to assess and develop driving dynamics for a low perceived vehicle weight of BEVs. The so-called lightweight feeling. In addition, measures for improvement, their impact as well as their related costs are introduced. Beside the characterization and identification of measures for improvement with AVLs attribute engineering approach, a methodology, which allows to resolve and optimize that the tension between costs, efficiency and driving experience very early in the virtual concept development, will be addressed.

The approval of highly automated (HAD) vehicles and modern advanced driver assistance systems (ADAS) requires extensive tests in order to guarantee safety and driving comfort in all driving situations. Adjustments of the vehicle settings that result from such tests during an ongoing approval process are associated with great effort. Changes and developments at the controller and actuator level in this phase are therefore more difficult or even require a restart of the whole test phase.With the aim of overcoming this obstacle in the area of chassis development, a tool for driving behavior-based safeguarding has been developed at TWT GmbH. The tool deployment itself was in close cooperation with and for our customers from the automotive industry. On the basis of numerical simulations, this development enables to detect the mutual influences of a component adjustment on the behavior of the entire vehicle. For this purpose, the three vehicle levels – tactical, controller and actuator – are considered separately and interfaces between the levels are precisely defined. A limited set of evaluation parameters and test scenarios can be derived, which enable simulation-based safeguarding for changes to the chassis controllers and actuators. The decisive factor in this approach is that the assessment of interactions does not take place on the microscopic scale of signal and effect chains, but is derived from the macroscopic domain of driving behavior.

Battery electric vehicles (BEVs) can represent a feasible solution for reaching the legislative CO2 reduction targets. Li-Ion batteries are the most promising candidates for BEVs due to their high energy density (200 Wh/kg). Heat generation during charge/discharge processes causes temperature increase and thermal management is indispensable. Also temperature gradients inside a cell and a pack must be minimized. This work investigates the thermal behaviour of battery cells during charging/discharging processes both experimentally and numerically. Measurements of several charging/discharging cycles at different ambient temperatures and C-rates are used to build mathematical models of the heat generation due to ohmic/kinetic resistances and entropy variation in function of temperature and state of charge. Numerical simulations in OpenFOAM solve the heat conduction problem inside the cell with air convection at the external surfaces and a heat source in the core. A parametric study to evaluate the impact of discharge/charge rate, ambient temperature, convection conditions and material properties is proposed. Heat losses at slow discharge rates (<0.5C) are below 1% of the power output of the cell and they depend similarly both from ohmic resistance and entropy variation, while at higher discharging rates total heat generation rates reach 5% (at 5C rates) of the power output. Moreover, the heat generation increases disproportionally at very low state of charge levels, SoC < 0.1.

This publication extends on a holistic system design approach by implementing an efficient simulation methodology to design a robust cooling system for a liquid cooled battery module. After gaining insight into the functionality of the system and its components, the most promising parts for redesign were selected using a method called Extended Target Weighing Approach (ETWA). Because of this redesign, the cell mount was reduced in height and the cooling system topology parallelized to lower pressure losses. The design of the new cooling system geometry is explored. The parameter optimization is carried out with a simplified 1D fluid dynamics model. To verify and validate the 1D-simulation, a more detailed 3D-simulation is used at bespoke crucial times during development. The main boundary conditions were keeping the pressure losses at or below its reference level from the module’s predecessor, while designing a robust system in terms of flow rate distribution in the parallel channels. The result was a system behaving almost indifferent to changes in temperature or total flow rate with the channels individually changing less than 5% of total flow rate and all channels staying in the range between 20% to 32% of total flow rate at all times. The overall system pressure drop could be brought down to half of its reference value. The two simulations were achieving similar results, eventually deviating less than 5% relative to each other in simulated flow rates and pressure losses.

In this paper, software algorithms for condition monitoring of a battery management system (BMS) are developed. The given application is a 400 V high-voltage (HV) battery-pack for automotive use. Generally, the BMS is responsible for a safe and optimized operation of the HV-battery-pack. Therefore, a Matlab/Simulink model for an HV-battery-pack is designed and the common WLTP-load profile computed. For the BMS a state-of-the-art Coulomb-Counter for battery State-of-Charge estimation is implemented. Furthermore, a three-criteria current limitation is realized to perform high quality power prediction. The battery aging-state based on the internal resistance $$SOH_{R}$$ S O H R is estimated through an internal resistance estimation using several methods. Notably, the internal resistance estimation during a dynamic driving cycle leads to promising results. During a WLTP driving cycle the RT2s-method is able to estimate the internal resistance with a $$3\sigma $$ 3 σ -accuracy of less than $$0{,}07\,m\Omega$$ 0 , 07 m Ω . The RT2s-method combined with an adaptive lowpass filter shows a high level of accuracy, stability and robustness. The implemented algorithm detects the $$SOH_{R}$$ S O H R at different settings with a maximum error of $$\Delta SOH_{R} <$$ Δ S O H R < 8 %.