The automotive industry has been funding warranty repair work for many decades. The most common vehicle warranty is 3 years or 36,000 miles [1]. Original equipment manufacturers (OEM) in North America have dealers record all the work completed and submit claims for the work that qualifies for warranty reimbursement [2]. The OEM reviews the request and pays dealers for the work performed. In addition to payments, the database is also used to complete quality analysis for the vehicles. Often the software being used by dealerships is old and not designed for quality analysis. Reviewing all the warranty work done can be an arduous task. OEMs can receive 100,000 or more claims each day. To speed up the analysis process the OEMs will divide the repair work into sections based on the segment of the vehicle requiring work. This categorization allows the OEMs to spread the work across many experts in the company. But what does the OEMs do when the problem cannot be located at the dealership? The dealer still requires payment for the time they spent trying to find the issue. This is often categorized as TNF (trouble not found). This type of work without a resolution can account for a sizable percentage of warranty costs! It can be as high as 38% depending on the manufacturer. It can also affect customer satisfaction and customer loyalty. In fact, one of the most common complaints submitted to NHTSA (National Highway Traffic Safety Administration) is “dealer unable to locate problem” [3]. To make matters worse, most customers end up returning to the dealer multiple times before the issue is, if ever, located, and fixed [4]. So how can we find a better way to analyze these warranty claims and improve the customer experience while decreasing the cost for the OEM? This problem can be improved through machine learning and statistical pattern analysis.
{"title":"Warranty Repairs Reimagined through Machine Learning and Statistical\u0000 Pattern Recognition (Part 1)","authors":"Jody Hand, Sawyer Hall, Michael Carr, Jeremy Worm","doi":"10.4271/2024-01-5071","DOIUrl":"https://doi.org/10.4271/2024-01-5071","url":null,"abstract":"The automotive industry has been funding warranty repair work for many decades.\u0000 The most common vehicle warranty is 3 years or 36,000 miles [1]. Original equipment manufacturers (OEM)\u0000 in North America have dealers record all the work completed and submit claims\u0000 for the work that qualifies for warranty reimbursement [2]. The OEM reviews the request and pays dealers for the\u0000 work performed. In addition to payments, the database is also used to complete\u0000 quality analysis for the vehicles. Often the software being used by dealerships\u0000 is old and not designed for quality analysis. Reviewing all the warranty work\u0000 done can be an arduous task. OEMs can receive 100,000 or more claims each day.\u0000 To speed up the analysis process the OEMs will divide the repair work into\u0000 sections based on the segment of the vehicle requiring work. This categorization\u0000 allows the OEMs to spread the work across many experts in the company. But what\u0000 does the OEMs do when the problem cannot be located at the dealership? The\u0000 dealer still requires payment for the time they spent trying to find the issue.\u0000 This is often categorized as TNF (trouble not found). This type of work without\u0000 a resolution can account for a sizable percentage of warranty costs! It can be\u0000 as high as 38% depending on the manufacturer. It can also affect customer\u0000 satisfaction and customer loyalty. In fact, one of the most common complaints\u0000 submitted to NHTSA (National Highway Traffic Safety Administration) is “dealer\u0000 unable to locate problem” [3]. To make\u0000 matters worse, most customers end up returning to the dealer multiple times\u0000 before the issue is, if ever, located, and fixed [4]. So how can we find a better way to analyze these warranty claims\u0000 and improve the customer experience while decreasing the cost for the OEM? This\u0000 problem can be improved through machine learning and statistical pattern\u0000 analysis.","PeriodicalId":510086,"journal":{"name":"SAE Technical Paper Series","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141826438","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Azim Turan, Emrah Yaki, Çağrı Emre Birgül, Hakan Kaya
Improving the overall thermal management strategy in electric vehicles can directly or indirectly improve battery efficiency and vehicle range. The aim of this study is to simulate and improve the performance of motor cooling lines developed for an electric bus using 1D computational fluid dynamics models. In the study, simulation studies were carried out for the 12-m EV Citivolt vehicle of Anadolu Isuzu. Design parameters such as placement of flow lines and component selection were decided thanks to 1D models developed. The design obtained at the end of the study was supported by tests and experimental studies. As a result, it was seen that the components in the line correctly detected the flow rate and pressure losses with a maximum error rate of 8% and an average error rate of 4.8%. Additionally, the components on the line were added to the model via their own characteristic dp-Q curves. In this way, it has been seen that these components, which contain complex flow lines, can be represented with unique dp-Q curves.
{"title":"Analysis of Motor Cooling Lines of Electric Buses with 1D Models and\u0000 Verification with Experiments","authors":"Azim Turan, Emrah Yaki, Çağrı Emre Birgül, Hakan Kaya","doi":"10.4271/2024-01-5070","DOIUrl":"https://doi.org/10.4271/2024-01-5070","url":null,"abstract":"Improving the overall thermal management strategy in electric vehicles can\u0000 directly or indirectly improve battery efficiency and vehicle range. The aim of\u0000 this study is to simulate and improve the performance of motor cooling lines\u0000 developed for an electric bus using 1D computational fluid dynamics models. In\u0000 the study, simulation studies were carried out for the 12-m EV Citivolt vehicle\u0000 of Anadolu Isuzu. Design parameters such as placement of flow lines and\u0000 component selection were decided thanks to 1D models developed. The design\u0000 obtained at the end of the study was supported by tests and experimental\u0000 studies. As a result, it was seen that the components in the line correctly\u0000 detected the flow rate and pressure losses with a maximum error rate of 8% and\u0000 an average error rate of 4.8%. Additionally, the components on the line were\u0000 added to the model via their own characteristic dp-Q curves. In this way, it has\u0000 been seen that these components, which contain complex flow lines, can be\u0000 represented with unique dp-Q curves.","PeriodicalId":510086,"journal":{"name":"SAE Technical Paper Series","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141826509","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Liquid jet atomization is one of the key processes in many engineering applications, such as IC engines, gas turbines, and the like, to name a few. Simulating this process using a pure Eulerian or a pure Lagrangian framework has its own drawbacks. The Eulerian–Lagrangian spray atomization (ELSA) modeling seems like a viable alternative in such scenarios. ELSA simulations consist of solving an additional transport equation for the surface area density (Σ) of the issuing jet. In this study we have proposed a dynamic approach to compute the turbulent timescale constant (α1), which appears in the source of Σ-transport equation and is responsible for restoring the surface area back to its equilibrium. The dynamic approach involves an analytical computation of the turbulent timescale constant (α1), thereby eliminating the need for ad hoc adjustments to surface area values during computational fluid dynamics (CFD) simulations. Unlike previous research which suggests using constant values in the range (0, 1] for the α1-constant, we found that these values can be as high as 60,000 for the engine combustion network (ECN) spray-A nozzle conditions. The analytical closure procedure dampens the spurious overshoots seen in the sigma-Y field and maintains values close to the equilibrium conditions. The proposed approach is implemented in CONVERGE, a commercially available CFD code and validated by comparing against available experimental data.
{"title":"In Situ Estimation of the Coefficient of Stress Source in the\u0000 Eulerian–Lagrangian Spray Atomization Model","authors":"C. R. L. Anumolu, Ambarish Dahale","doi":"10.4271/2024-01-5069","DOIUrl":"https://doi.org/10.4271/2024-01-5069","url":null,"abstract":"Liquid jet atomization is one of the key processes in many engineering\u0000 applications, such as IC engines, gas turbines, and the like, to name a few.\u0000 Simulating this process using a pure Eulerian or a pure Lagrangian framework has\u0000 its own drawbacks. The Eulerian–Lagrangian spray atomization (ELSA) modeling\u0000 seems like a viable alternative in such scenarios. ELSA simulations consist of\u0000 solving an additional transport equation for the surface area density (Σ) of the\u0000 issuing jet. In this study we have proposed a dynamic approach to compute the\u0000 turbulent timescale constant (α1), which appears in\u0000 the source of Σ-transport equation and is responsible for restoring the surface\u0000 area back to its equilibrium. The dynamic approach involves an analytical\u0000 computation of the turbulent timescale constant\u0000 (α1), thereby eliminating the need for ad hoc\u0000 adjustments to surface area values during computational fluid dynamics (CFD)\u0000 simulations. Unlike previous research which suggests using constant values in\u0000 the range (0, 1] for the α1-constant, we found that\u0000 these values can be as high as 60,000 for the engine combustion network (ECN)\u0000 spray-A nozzle conditions. The analytical closure procedure dampens the spurious\u0000 overshoots seen in the sigma-Y field and maintains values close to the\u0000 equilibrium conditions. The proposed approach is implemented in CONVERGE, a\u0000 commercially available CFD code and validated by comparing against available\u0000 experimental data.","PeriodicalId":510086,"journal":{"name":"SAE Technical Paper Series","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141681908","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The Single Cylinder Research Engine (SCRE) at the Institute of Internal Combustion Engines and Powertrain Systems is equipped with a variable valve train that allows to switch between regular intake valve lift and early intake valve closing (Miller). On the exhaust side, a secondary exhaust valve lift (SEVL) on each valve is possible with adjustable back pressure and thus the possibility of realizing internal EGR. In combination with alternative fuels, even if they are Drop-In capable as HVO, properties differ and can influence the emission and efficiency behavior. The investigations of this paper are focusing on regenerative Drop-In fuel (HVO), fossil fuel (B7), and an oxygenate (OME), that needs adaptions at the engine control unit, but offers further emission potential. By commissioning a 2-stage boost system, it is possible to fully equalize the air mass in Miller mode compared to the normal valve lift. This enables a comprehensive analysis of the behavior of the fuels under different boundary conditions. In addition to the boost pressure, the exhaust gas pressure and engine speed are varied and analyzed with regards to emissions and efficiency. The SEVL is varied and investigated in terms of emission and efficiency behavior. For the evaluation, a combustion analysis is carried out and analyzed based on cylinder pressure data to work out the causes of the respective effects. One expected effect is a NOx reduction in Miller mode with the same air mass due to reduced effective compression, without significant efficiency losses due to the constant expansion. In the investigations this effect is clearly visible and therefore represents great potential for reducing NOx emissions.
{"title":"Miller Cycle and Internal EGR in Diesel Engines Using Alternative Fuels","authors":"Friedemar Knost, Christian Beidl","doi":"10.4271/2024-01-3020","DOIUrl":"https://doi.org/10.4271/2024-01-3020","url":null,"abstract":"The Single Cylinder Research Engine (SCRE) at the Institute of Internal Combustion Engines and Powertrain Systems is equipped with a variable valve train that allows to switch between regular intake valve lift and early intake valve closing (Miller). On the exhaust side, a secondary exhaust valve lift (SEVL) on each valve is possible with adjustable back pressure and thus the possibility of realizing internal EGR. In combination with alternative fuels, even if they are Drop-In capable as HVO, properties differ and can influence the emission and efficiency behavior. The investigations of this paper are focusing on regenerative Drop-In fuel (HVO), fossil fuel (B7), and an oxygenate (OME), that needs adaptions at the engine control unit, but offers further emission potential. By commissioning a 2-stage boost system, it is possible to fully equalize the air mass in Miller mode compared to the normal valve lift. This enables a comprehensive analysis of the behavior of the fuels under different boundary conditions. In addition to the boost pressure, the exhaust gas pressure and engine speed are varied and analyzed with regards to emissions and efficiency. The SEVL is varied and investigated in terms of emission and efficiency behavior. For the evaluation, a combustion analysis is carried out and analyzed based on cylinder pressure data to work out the causes of the respective effects. One expected effect is a NOx reduction in Miller mode with the same air mass due to reduced effective compression, without significant efficiency losses due to the constant expansion. In the investigations this effect is clearly visible and therefore represents great potential for reducing NOx emissions.","PeriodicalId":510086,"journal":{"name":"SAE Technical Paper Series","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141686227","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Society is moving towards climate neutrality where hydrogen fuelled combustion engines (H2 ICE) could be considered a main technology. These engines run on hydrogen (H2) so carbon-based emission are only present at a very low level from the lube oil. The most important pollutants NO and NO2 are caused by the exhaust aftertreatment system as well as CO2 coming from the ambient air.
For standard measurement technologies these low levels of CO2 are hard to detect due to the high-water content. Normal levels of CO2 are between 400-500 ppm which is very close or even below the detection limit of commonly used non-dispersive-infrared-detectors (NDIR). As well the high-water content is very challenging for NOx measuring devices, like chemiluminescence detectors (CLD), where it results in higher noise and therefore a worse detection limit. Even for Fourier-transformed-infrared-spectroscopy-analysers (FT-IR) it is challenging to deal with water content over 15% without increased noise.
The goal of this study was to show that measuring low levels of CO2 and NOx can be performed by FT-IR. Therefore, new calibrations are created for NO, NO2 and CO2. These were first tested by calibration gases in wet and dry conditions. Afterwards a 2l four-cylinder passenger car engine that is run on hydrogen was used to generate real engine data. The engine is equipped with a state-of-the-art SCR catalyst system. The FT-IR analyser was compared to theoretical data as well as to a standard CLD and NDIR. Several steady-state points were performed as well as a driving cycle.
The results show a large improvement in reducing the noise caused by high water and therefore a more accurate measurement at low concentrations. Measured concentrations as well as masses show a good alignment with expected values.
社会正朝着气候中和的方向发展，氢燃料内燃机（H2 ICE）可被视为一种主要技术。这些发动机以氢（H2）为燃料，因此只有极少量的碳排放来自润滑油。最重要的污染物 NO 和 NO2 由排气后处理系统以及来自环境空气的 CO2 引起。二氧化碳的正常含量在 400-500 ppm 之间，非常接近甚至低于常用的非色散红外探测器 (NDIR) 的检测极限。此外，高含水量对于氮氧化物测量设备（如化学发光检测器 (CLD)）来说也是非常具有挑战性的，它会导致更高的噪声，从而降低检测限。即使是傅立叶变换红外光谱分析仪（FT-IR），要在不增加噪声的情况下处理含水量超过 15%的问题也很有挑战性。这项研究的目的是要证明，可以用 FT-IR 测量低浓度的二氧化碳和氮氧化物。因此，为 NO、NO2 和 CO2 创建了新的定标。首先在潮湿和干燥条件下用校准气体对其进行了测试。然后，使用一台以氢为燃料的 2 升四缸乘用车发动机生成真实的发动机数据。该发动机配备了最先进的 SCR 催化剂系统。傅立叶变换红外分析仪与理论数据以及标准 CLD 和 NDIR 进行了比较。结果表明，该分析仪在减少高浓度水引起的噪音方面有很大改进，因此在低浓度时测量更加精确。测量到的浓度和质量与预期值非常吻合。
{"title":"Challenges of Measuring Low Levels of CO\u00002\u0000 and NO\u0000x\u0000 on H\u00002\u0000-ICE","authors":"Philipp Jakubec, Sebastian Roiser","doi":"10.4271/2024-01-2998","DOIUrl":"https://doi.org/10.4271/2024-01-2998","url":null,"abstract":"<div class=\"section abstract\"><div class=\"htmlview paragraph\">Society is moving towards climate neutrality where hydrogen fuelled combustion engines (H<sub>2</sub> ICE) could be considered a main technology. These engines run on hydrogen (H<sub>2</sub>) so carbon-based emission are only present at a very low level from the lube oil. The most important pollutants NO and NO<sub>2</sub> are caused by the exhaust aftertreatment system as well as CO<sub>2</sub> coming from the ambient air.</div><div class=\"htmlview paragraph\">For standard measurement technologies these low levels of CO<sub>2</sub> are hard to detect due to the high-water content. Normal levels of CO<sub>2</sub> are between 400-500 ppm which is very close or even below the detection limit of commonly used non-dispersive-infrared-detectors (NDIR). As well the high-water content is very challenging for NO<sub>x</sub> measuring devices, like chemiluminescence detectors (CLD), where it results in higher noise and therefore a worse detection limit. Even for Fourier-transformed-infrared-spectroscopy-analysers (FT-IR) it is challenging to deal with water content over 15% without increased noise.</div><div class=\"htmlview paragraph\">The goal of this study was to show that measuring low levels of CO<sub>2</sub> and NO<sub>x</sub> can be performed by FT-IR. Therefore, new calibrations are created for NO, NO<sub>2</sub> and CO<sub>2</sub>. These were first tested by calibration gases in wet and dry conditions. Afterwards a 2l four-cylinder passenger car engine that is run on hydrogen was used to generate real engine data. The engine is equipped with a state-of-the-art SCR catalyst system. The FT-IR analyser was compared to theoretical data as well as to a standard CLD and NDIR. Several steady-state points were performed as well as a driving cycle.</div><div class=\"htmlview paragraph\">The results show a large improvement in reducing the noise caused by high water and therefore a more accurate measurement at low concentrations. Measured concentrations as well as masses show a good alignment with expected values.</div></div>","PeriodicalId":510086,"journal":{"name":"SAE Technical Paper Series","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141687169","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The pace of innovations in battery development is revolutionizing the landscape and opportunities for energy storage applications leading to a stronger market segmentation enabling a better suitability to fulfill specific application requirements.For automotive applications, several approaches to increase energy densities, to improve fast charging performance, and to reduce cost on a pack level are considered. Among them, a promising example is the direct integration of battery cells into the battery pack (Cell-to-pack; CTP) or vehicle (Cell-to-chassis, CTC) to increase energy densities and to reduce costs, as already commercialized by Tesla, CATL and others.On cell level, a segmentation between high-performance and low-cost applications is realized in the technology developments. Hereby, a diversification of the cell manufacturer’s product portfolio can be observed. As a strong demand for NMC and LFP-based battery cells is leading to fluctuating raw material prices (especially for Lithium), especially sodium-ion batteries (SiB) are gaining increased attention further pushed with the announcements and achievements of CATL and Northvolt to be considered primarily for ESS, but also potentially automotive applications. On the other hand, developments to improve cell performance are achieving higher maturity, especially on the anode side with the implementation of engineered silicon materials, but also for (semi-)solid-state technology. As a compromise, mixed LF(M)P/NMC cell chemistry concepts are also considered as alternative solutions.
{"title":"Next-Gen Battery Strategies 2027+: Potentials and Challenges for Future Battery Designs and Diversification in Product Portfolios to Serve a Large Bandwidth of Market Applications","authors":"Ines Miller","doi":"10.4271/2024-01-3018","DOIUrl":"https://doi.org/10.4271/2024-01-3018","url":null,"abstract":"The pace of innovations in battery development is revolutionizing the landscape and opportunities for energy storage applications leading to a stronger market segmentation enabling a better suitability to fulfill specific application requirements.For automotive applications, several approaches to increase energy densities, to improve fast charging performance, and to reduce cost on a pack level are considered. Among them, a promising example is the direct integration of battery cells into the battery pack (Cell-to-pack; CTP) or vehicle (Cell-to-chassis, CTC) to increase energy densities and to reduce costs, as already commercialized by Tesla, CATL and others.On cell level, a segmentation between high-performance and low-cost applications is realized in the technology developments. Hereby, a diversification of the cell manufacturer’s product portfolio can be observed. As a strong demand for NMC and LFP-based battery cells is leading to fluctuating raw material prices (especially for Lithium), especially sodium-ion batteries (SiB) are gaining increased attention further pushed with the announcements and achievements of CATL and Northvolt to be considered primarily for ESS, but also potentially automotive applications. On the other hand, developments to improve cell performance are achieving higher maturity, especially on the anode side with the implementation of engineered silicon materials, but also for (semi-)solid-state technology. As a compromise, mixed LF(M)P/NMC cell chemistry concepts are also considered as alternative solutions.","PeriodicalId":510086,"journal":{"name":"SAE Technical Paper Series","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141686317","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Dejie Ji, Maximilian Flormann, Joana M. Warnecke, Roman Henze, T. Deserno
Investigating human driver behavior enhances the acceptance of the autonomous driving and increases road safety in heterogeneous environments with human-operated and autonomous vehicles. The previously established driver fingerprint model, focuses on the classification of driving styles based on CAN bus signals. However, driving styles are inherently complex and influenced by multiple factors, including changing driving environments and driver states. To comprehensively create a driver profile, an in-car measurement system based on the Driver-Driven vehicle-Driving environment (3D) framework is developed. The measurement system records emotional and physiological signals from the driver, including the ECG signal and heart rate. A Raspberry Pi camera is utilized on the dashboard to capture the driver's facial expressions and a trained convolutional neural network (CNN) recognizes emotion. To conduct unobtrusive ECG measurements, an ECG sensor is integrated into the steering wheel. Additionally, the system accesses CAN bus signals from the vehicle to assess the driver’s driving style, extracting signals related to longitudinal and lateral control behavior from the Drive-CAN (A-CAN). Recognizing that variables from the driving environment can influence driving style, such as traffic signs and road conditions, a windshield-mounted webcam is integrated into the measurement system. This setup enables real-time detection of common traffic signs and assessment of road conditions, distinguishing between dry, wet, or icy road surfaces. Augmenting of the image data from camera, signals from in-car ADAS-sensors, such as the distance measured by the front radar in relation to neighboring vehicles, are integrated for a comprehensive analysis of driving style. The established measurement system is presently implemented in a test vehicle, poised to investigate the interplay between the 3D-parameters, with a focus on driving style of human driver.
对人类驾驶员行为的研究可提高人们对自动驾驶的接受程度，并在有人类驾驶车辆和自动驾驶车辆的异构环境中提高道路安全性。之前建立的驾驶员指纹识别模型侧重于根据 CAN 总线信号对驾驶风格进行分类。然而，驾驶风格本身十分复杂，并受到多种因素的影响，包括不断变化的驾驶环境和驾驶员状态。为了全面建立驾驶员特征，我们开发了基于驾驶员驱动车辆-驾驶环境（3D）框架的车载测量系统。该测量系统可记录驾驶员的情绪和生理信号，包括心电图信号和心率。仪表盘上的树莓派（Raspberry Pi）摄像头用于捕捉驾驶员的面部表情，训练有素的卷积神经网络（CNN）可识别情绪。为了进行不显眼的心电图测量，方向盘上集成了一个心电图传感器。此外，该系统还能访问来自车辆的 CAN 总线信号，以评估驾驶员的驾驶风格，并从 Drive-CAN (A-CAN) 中提取与纵向和横向控制行为相关的信号。由于认识到驾驶环境中的变量会影响驾驶风格，例如交通标志和路况，因此测量系统中集成了一个安装在挡风玻璃上的网络摄像头。这种设置可以实时检测常见的交通标志和评估路况，区分干燥、潮湿或结冰的路面。除了摄像头提供的图像数据外，车载 ADAS 传感器（如前雷达测得的与邻近车辆的距离）提供的信号也被整合进来，以便对驾驶风格进行综合分析。已建立的测量系统目前正在测试车辆中实施，准备研究三维参数之间的相互作用，重点是人类驾驶员的驾驶风格。
{"title":"Set-Up of an In-Car System for Investigating Driving Style on the Basis of the 3D-Method","authors":"Dejie Ji, Maximilian Flormann, Joana M. Warnecke, Roman Henze, T. Deserno","doi":"10.4271/2024-01-3001","DOIUrl":"https://doi.org/10.4271/2024-01-3001","url":null,"abstract":"Investigating human driver behavior enhances the acceptance of the autonomous driving and increases road safety in heterogeneous environments with human-operated and autonomous vehicles. The previously established driver fingerprint model, focuses on the classification of driving styles based on CAN bus signals. However, driving styles are inherently complex and influenced by multiple factors, including changing driving environments and driver states. To comprehensively create a driver profile, an in-car measurement system based on the Driver-Driven vehicle-Driving environment (3D) framework is developed. The measurement system records emotional and physiological signals from the driver, including the ECG signal and heart rate. A Raspberry Pi camera is utilized on the dashboard to capture the driver's facial expressions and a trained convolutional neural network (CNN) recognizes emotion. To conduct unobtrusive ECG measurements, an ECG sensor is integrated into the steering wheel. Additionally, the system accesses CAN bus signals from the vehicle to assess the driver’s driving style, extracting signals related to longitudinal and lateral control behavior from the Drive-CAN (A-CAN). Recognizing that variables from the driving environment can influence driving style, such as traffic signs and road conditions, a windshield-mounted webcam is integrated into the measurement system. This setup enables real-time detection of common traffic signs and assessment of road conditions, distinguishing between dry, wet, or icy road surfaces. Augmenting of the image data from camera, signals from in-car ADAS-sensors, such as the distance measured by the front radar in relation to neighboring vehicles, are integrated for a comprehensive analysis of driving style. The established measurement system is presently implemented in a test vehicle, poised to investigate the interplay between the 3D-parameters, with a focus on driving style of human driver.","PeriodicalId":510086,"journal":{"name":"SAE Technical Paper Series","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141686187","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jan Kramer, Michael Rice, Bryan Zavala, Christopher Sharp, James McCarthy, Ben Karrer
Engine and aftertreatment solutions have been identified to meet the upcoming ultra-low NOx regulations on heavy duty vehicles in the United States. These standards will require changes to current conventional aftertreatment systems for dealing with low exhaust temperature scenarios while increasing the useful life of the engine and aftertreatment system. Previous studies have shown feasibility of meeting the US EPA and California Air Resource Board (CARB) requirements. This work includes a 15L diesel engine equipped with cylinder deactivation (CDA) and an aftertreatment system that was fully DAAAC aged to 800,000 miles. The aftertreatment system includes an e-heater (electric heater), light-off Selective Catalytic Reduction (LO-SCR) followed by a primary aftertreatment system containing a DPF and SCR. The e-heater was capable of providing up to 10 kW, however for the purpose of this project, lower power settings of 2.5 kW and 5 kW were studied in combination with CDA for lowest possible CO2 emissions. Test cycles included the heavy duty FPT (hot and cold), a low load cycle, a beverage cycle, and a stay hot operation. The data found in this study show how the application of a low power e-heater enables an engine with CDA and an aftertreatment system after 800,000 miles aging to meet ultra-low NOx emissions with minimum CO2 penalty.
{"title":"Future Emissions Following 800,000 Mile Aging Using CDA and a Low Power Electric Heater","authors":"Jan Kramer, Michael Rice, Bryan Zavala, Christopher Sharp, James McCarthy, Ben Karrer","doi":"10.4271/2024-01-3011","DOIUrl":"https://doi.org/10.4271/2024-01-3011","url":null,"abstract":"Engine and aftertreatment solutions have been identified to meet the upcoming ultra-low NOx regulations on heavy duty vehicles in the United States. These standards will require changes to current conventional aftertreatment systems for dealing with low exhaust temperature scenarios while increasing the useful life of the engine and aftertreatment system. Previous studies have shown feasibility of meeting the US EPA and California Air Resource Board (CARB) requirements. This work includes a 15L diesel engine equipped with cylinder deactivation (CDA) and an aftertreatment system that was fully DAAAC aged to 800,000 miles. The aftertreatment system includes an e-heater (electric heater), light-off Selective Catalytic Reduction (LO-SCR) followed by a primary aftertreatment system containing a DPF and SCR. The e-heater was capable of providing up to 10 kW, however for the purpose of this project, lower power settings of 2.5 kW and 5 kW were studied in combination with CDA for lowest possible CO2 emissions. Test cycles included the heavy duty FPT (hot and cold), a low load cycle, a beverage cycle, and a stay hot operation. The data found in this study show how the application of a low power e-heater enables an engine with CDA and an aftertreatment system after 800,000 miles aging to meet ultra-low NOx emissions with minimum CO2 penalty.","PeriodicalId":510086,"journal":{"name":"SAE Technical Paper Series","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141688135","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Typically, machine learning techniques are used to realise autonomous driving. Be it as part of environment recognition or ultimately when making driving decisions. Machine learning generally involves the use of stochastic methods to provide statistical inference. Failures and wrong decisions are unavoidable due to the statistical nature of machine learning and are often directly related to root causes that cannot be easily eliminated. The quality of these systems is normally indicated by statistical indicators such as accuracy and precision. Providing evidence that accuracy and precision of these systems are sufficient to guarantee a safe operation is key for the acceptance of autonomous driving. Usually, tests and simulations are extensively used to provide this kind of evidence.However, the basis of all descriptive statistics is a random selection from a probability space. A major challenge in testing or constructing the training and test data set is that this probability space is usually not well defined. To systematically address this shortcoming, ontologies have been and are being developed to capture the various concepts and properties of the operational design domain. They serve as a basis for the specification of appropriate tests in different approaches. However, in order to make statistical statements about the system, information about the realistic frequency of the inferred test cases is still missing. Related to this problem is the proof of completeness and balance of the training data. While an ontology may be able to check the completeness of the training data, it lacks any information to prove its representativeness. In this article, we propose the extension of ontologies to include probabilistic information. This allows to evaluate the completeness and balance of training sets. Moreover, it serves as a basis for a random sampling of test cases, which allows mathematically sound statistical proofs of the quality of the ML system. We demonstrate our approach by extending published ontologies that capture typical scenarios of autonomous driving systems with probabilistic information.
{"title":"Probabilistically Extended Ontologies: A Basis for Systematic Testing of ML-Based Systems","authors":"H. Wiesbrock, Jürgen Grossmann","doi":"10.4271/2024-01-3002","DOIUrl":"https://doi.org/10.4271/2024-01-3002","url":null,"abstract":"Typically, machine learning techniques are used to realise autonomous driving. Be it as part of environment recognition or ultimately when making driving decisions. Machine learning generally involves the use of stochastic methods to provide statistical inference. Failures and wrong decisions are unavoidable due to the statistical nature of machine learning and are often directly related to root causes that cannot be easily eliminated. The quality of these systems is normally indicated by statistical indicators such as accuracy and precision. Providing evidence that accuracy and precision of these systems are sufficient to guarantee a safe operation is key for the acceptance of autonomous driving. Usually, tests and simulations are extensively used to provide this kind of evidence.However, the basis of all descriptive statistics is a random selection from a probability space. A major challenge in testing or constructing the training and test data set is that this probability space is usually not well defined. To systematically address this shortcoming, ontologies have been and are being developed to capture the various concepts and properties of the operational design domain. They serve as a basis for the specification of appropriate tests in different approaches. However, in order to make statistical statements about the system, information about the realistic frequency of the inferred test cases is still missing. Related to this problem is the proof of completeness and balance of the training data. While an ontology may be able to check the completeness of the training data, it lacks any information to prove its representativeness. In this article, we propose the extension of ontologies to include probabilistic information. This allows to evaluate the completeness and balance of training sets. Moreover, it serves as a basis for a random sampling of test cases, which allows mathematically sound statistical proofs of the quality of the ML system. We demonstrate our approach by extending published ontologies that capture typical scenarios of autonomous driving systems with probabilistic information.","PeriodicalId":510086,"journal":{"name":"SAE Technical Paper Series","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141685033","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Fabian Weitz, C. Debnar, Michael Frey, Frank Gauterin
Society's growing environmental awareness and increasing urbanisation require new and innovative vehicle concepts. The use of additive manufacturing (AM) expands the design freedom in component development. In this paper, these are utilised to further develop a front axle suspension for a new type of modular vehicle concept. The wheel suspension components are optimised on the basis of a new method that has already been applied in previous work. This is based on industry-standard load cases for the strength design of the components, as well as the available installation space determined for the design of the suspension components and the suitable configuration of the suspension components. The component geometries identified using numerical methods that are suitable for the force flow are optimised with regard to the integration of information, energy and material-carrying lines in the control arms and the lines are used as load-bearing structures as extensively as possible. High-strength light metals are used to minimise the component masses. Openings are provided in the components for routing electrical cables. The fluid transport is realised using lines integrated into the wishbones. The final geometries of the suspension components are then validated by a finite element analysis (FEA) of the entire suspension model. The result of the method used are lighter suspension components with a maximum degree of functional integration. The increased functional integration reduces the required installation space, which improves the vehicle package and achieves greater front wheel clearance, which increases the possible steering angles and thus improves maneuverability. The reduction in unsprung masses can improve driving behaviour and has a positive effect on the vehicle's energy consumption. In addition, the integration of the conductions section simplifies the assembly of the front axle suspension.
{"title":"Additively Manufactured Wheel Suspension System with Integrated Conductions and Optimized Structure","authors":"Fabian Weitz, C. Debnar, Michael Frey, Frank Gauterin","doi":"10.4271/2024-01-2973","DOIUrl":"https://doi.org/10.4271/2024-01-2973","url":null,"abstract":"Society's growing environmental awareness and increasing urbanisation require new and innovative vehicle concepts. The use of additive manufacturing (AM) expands the design freedom in component development. In this paper, these are utilised to further develop a front axle suspension for a new type of modular vehicle concept. The wheel suspension components are optimised on the basis of a new method that has already been applied in previous work. This is based on industry-standard load cases for the strength design of the components, as well as the available installation space determined for the design of the suspension components and the suitable configuration of the suspension components. The component geometries identified using numerical methods that are suitable for the force flow are optimised with regard to the integration of information, energy and material-carrying lines in the control arms and the lines are used as load-bearing structures as extensively as possible. High-strength light metals are used to minimise the component masses. Openings are provided in the components for routing electrical cables. The fluid transport is realised using lines integrated into the wishbones. The final geometries of the suspension components are then validated by a finite element analysis (FEA) of the entire suspension model. The result of the method used are lighter suspension components with a maximum degree of functional integration. The increased functional integration reduces the required installation space, which improves the vehicle package and achieves greater front wheel clearance, which increases the possible steering angles and thus improves maneuverability. The reduction in unsprung masses can improve driving behaviour and has a positive effect on the vehicle's energy consumption. In addition, the integration of the conductions section simplifies the assembly of the front axle suspension.","PeriodicalId":510086,"journal":{"name":"SAE Technical Paper Series","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141685894","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：