Search Results

In a recent time, which new vehicle lines comes with a huge number of sensors, control units, embedded technologies, and the complexity of these systems (electronics, electrical and electromechanical parts) increases in an exponential way. Considering these events, the expressive generated data amount grows in the same pace, so, consume, transform, and analyze all these data to better understand the modern customer, their needs and how they use the car features becomes necessary. Through that scenario, connected vehicles developed by Ford Motor Company has been generating opportunities to feature’s improvement and cost reduction based on data analysis. This growing quantity of data might be used to optimize feature systems and help engineering teams to understand how the features have been used and enhance the systems engineering design for new or existing features.

In recent years, the automotive industry has seen an exponential increase in the replacement of mechanical components with electronic-controlled components or systems. engine, transmission, brake, exhaust gas recirculation (EGR), lighting, driver-assist technologies, etc. are all monitored and/or controlled electronically. Connected vehicles are increasingly being used by Original Equipment Manufacturers (OEMs) to collect and transmit vehicle data in real-time via the use of various sensors, actuators, and communication technologies. Vehicle telematics devices can collect and transmit data about the vehicle location, speed, fuel efficiency, State Of Charge (SOC), auxiliary battery voltage, emissions, performance, and more. This data is sent over to the cloud via cellular networks, where it can be processed and analyzed to improve their products and services by automotive companies and/or fleet management.

To enable smooth and low-risk autonomous driving in the presence of other road users, such as cyclists and pedestrians, appropriate predictive safe speed control strategies relying on accurate and robust prediction models should be employed. However, difficulties related to driving scene understanding and a wide variety of features influencing decisions of other road users significantly complexifies prediction tasks and related controls. This paper proposes a hierarchical neural network (NN)-based prediction model of pedestrian crossing behavior, which is aimed to be applied within an autonomous vehicle (AV) safe speed control strategy. Additionally, different single-level prediction models are presented and analyzed as well, to serve as baseline approaches.

The Ford Motor Company Rolling Road Wind Tunnel (RRWT) is a state-of-the-art aerodynamic wind tunnel test facility in Allen Park, Michigan. The RRWT has operated since January 2022 and is designed for passenger and motorsport vehicle development. The test facility includes an office area, three secure customer vehicle preparation bays, a garage area, a vehicle frontal area measurement system, and a full-scale ¾ open jet wind tunnel. The wind tunnel features an interchangeable single belt and 5-belt Moving Ground Plane (MGP) system with an integrated 6-component balance, a two-position nozzle, boundary layer removal systems, and two independent flow traverse systems. Each flow traverse has a large horizontal box beam and vertical Z-strut that can position the flow traverse accurately within the test volume.

In modern automobiles, many new complex features are enabled by software and sensors. When combined with the variability of real-world environments and scenarios, validation of this ever-increasing amount of software becomes complex, costly, and takes a lot of time. This challenges automakers ability to quickly and reliably develop and deploy new features and experiences that their customers want in the marketplace. While traditional validation methods and modern virtual validation environments can cover most new feature testing, it is challenging to cover certain real-world scenarios. These scenarios include variation in weather conditions, roadway environments, driver usage, and complex vehicle interactions. The current approach to covering these scenarios often relies on data collected from long vehicle test trips that try to capture as many of these unique situations as possible. These test trips contribute significantly to the validation cost and time of new features.

To help address the issue of message authentication on the Controller Area Network (CAN) bus, researchers at Virginia Tech and Ford Motor Company have developed a proof-of-concept time-evolving watermark-based authentication mechanism that offers robust, cryptographically controlled confirmation of a CAN message's authenticity. This watermark is injected as a common-mode signal on both CAN-HI and CAN-LO bus voltages and has been proven using a low-cost software-defined radio (SDR) testbed. This paper extends prior analysis on the design and proof-of-concept to consider robustness testing over the range of voltages, both steady state drifts and transients, as are commonly witnessed within a vehicle. Overall performance results, along with a dynamic watermark amplitude control, validate the concept as being a practical near-term approach at improving authentication confidence of messages on the CAN bus.

Traditionally, the controls system in production vehicles with automatic transmission interprets the driver’s accelerator pedal position as a demand for transmission input torque. However, with the advent of electrified vehicles, where actuators are located at different positions in the drivetrain, and of autonomous vehicles, which are self-driving, it is more convenient to interpret the demand (either human or virtual) in vehicle acceleration or wheel torque domain. To this end, a Wheel Torque-based longitudinal Control (WTC) framework was developed, wherein demands can be converted accurately between the vehicle acceleration or wheel torque domain and the transmission assembly input torque domain.

Development of propulsion control systems frequently involves large-scale transient simulations, e.g. Monte Carlo simulations or drive-cycle optimizations, which require fast dynamic plant models. Models of the air-path—for internal combustion engines or fuel cells—can exhibit stiff behavior, though, causing slow numerical simulations due to either using an implicit solver or sampling much faster than the bandwidth of interest to maintain stability. This paper proposes a method to reduce air-path model stiffness by adding an impedance in series with potentially stiff components, e.g. throttles, valves, compressors, and turbines, thereby allowing the use of a fast-explicit solver. An impedance, by electrical analogy, is a frequency-dependent resistance to flow, which is shaped to suppress the high-frequency dynamics causing air-path stiffness, while maintaining model accuracy in the bandwidth of interest.

Multiple approaches have been created to enhance intra-vehicle communications security over the past three decades since the introduction of the Controller Area Network (CAN) protocol. The twin pair differential-mode communications bus is tremendously robust in the face of interference, yet physical access to the bus offers a variety of potential attack vectors whereby false messages and/or denial of service are achievable. This paper evaluates extensions of a Physical-layer (PHY) common-mode watermark-based authentication technique recently developed to improve authentication on the CAN bus by considering the watermark as a side-channel communications means for high value information. We also propose and analyze higher layer algorithms, with benefits and pitfalls, for employing the watermark as a physical-layer firewall.

As electric vehicles are becoming increasingly popular and necessary for the future mobility needs of civilization, further effort is continually made to improve the efficiency, cost, and safety of the lithium-ion battery packs that power these vehicles. To facilitate these goals, this paper introduces a data driven model to predict a distribution of surface temperatures for a lithium-ion battery pack: a long short-term memory (LSTM) neural network. The LSTM model is trained and validated with lithium-ion cells electrically connected to form a battery pack. Voltage, current, state of charge (SOC), and cell surface temperature from two arrays are used as inputs from a wide range of high and low temperature drive cycles. Additionally, ambient temperature is added as an input to the LSTM model.

In this paper, we present an end-to-end real-time detection and collision avoidance framework in an autonomous vehicle using a monocular RGB camera. The proposed system is able to run on embedded hardware in the vehicle to perform real-time detection of small objects. RetinaNet architecture with ResNet50 backbone is used to develop the object detection model using RGB images. A quantized version of the object detection inference model is implemented in the vehicle using NVIDIA Jetson AGX Xavier. A geometric method is used to estimate the distance to the detected object which is forwarded to a MicroAutoBox device that implements the control system of the vehicle and is responsible for maneuvering around the detected objects. The pipeline is implemented on a passenger vehicle and demonstrated in challenging conditions using different obstacles on a predefined set of waypoints.

In this paper, a novel methodology of nonlinear control is used, and a passivity-based control of contractive port-controlled Hamiltonian (PCH) systems is applied to a permanent magnet synchronous motor (PMSM). This methodology, also called “tIDA-PBC” (Trajectory Injection and Damping Assignment—Passivity-Based Control), uses passivity-based control of PCH systems “IDA-PBC” and exploits the properties of contractive Hamiltonian systems, resulting in a closed loop with its contractive system desired dynamics, thus obtaining an exponential trajectory tracking without relying on the error coordinates. In this system, a few steps are proposed in order to divide and modularize the methodology so it can be redesigned or reapplied in other systems by the reader. First, we define the model and set the way to solve the “matching equation.” Then the feasible and reference trajectories are obtained.

Researchers and engineers are utilizing big data analytics to draw further insights into transportation systems. Large amounts of data at the individual vehicle trip level are being collected and stored. The true potential of such data is still to be determined. In this paper, we are presenting a data-driven, novel, and intuitive approach to model driver behaviors using microscopic traffic simulation. Our approach utilizes metaheuristic methods to create an analytical tool to assess vehicle performance. Secondly, we show how microscopic simulation run outputs can be post-processed to obtain vehicle and trip level performance metrics. The methodology will form the basis for a data-driven approach to unearthing trip experiences as realized by drivers in the real world. The methodology will contribute to, A.) Using vehicle trajectory traces to identify underlying vehicle maneuver distributions as obtained from real-world driver data, B.)

A cast magnesium AE44 subframe was designed and manufactured for a C Class sedan to reduce weight and improve vehicle fuel economy. Corrosion mitigation strategies were developed to reduce the likelihood of galvanic corrosion. Both a proving ground vehicle corrosion test and a laboratory component corrosion test were conducted. The vehicle test result demonstrated that the corrosion mitigation strategies were effective. They also provided lessons learned on clearance between magnesium and steel components and options to improve the subframe’s corrosion resistance. The magnesium subframe achieved 5 kg (32%) weight reduction from the equivalent steel subframe and met all the required structural performance targets.

As the world searches for ways to reduce humanity’s impact on the environment, the automotive industry looks to extend the viable use of the gasoline engine by improving efficiency. One way to improve engine efficiency is through more effective control. Torque-based control is critical in modern cars and trucks for traction control, stability control, advanced driver assistance systems, and autonomous vehicle systems. Closed loop torque-based engine control systems require feedback signal(s); indicated mean effective pressure (IMEP) is a useful signal but is costly to measure directly with in-cylinder pressure sensors. Previous work has been done in torque and IMEP estimation using crankshaft acceleration and ion sensors, but these systems lack accuracy in some operating ranges and the ability to estimate cycle-cycle variation.

The software gradually takes over more and more tasks of the driver and paves the way to autonomous driving. Software development and software verification is therefore crucial for manufacturer's success. Standards such as ISO 26262 highly recommend requirements-based verification. Agile development uses continuous integration testing based on test automation and evaluation. All this pushed the creation of a model-based software verification environment that provides test generation and test automatization for all kinds of signal-based tests along the V-model. This paper presents a novel core component of this environment, which is as far as to the extent possible a standard-compliant cloud-based solution to test automation at the hardware level. Based on characteristic properties of testbenches, such as the wiring or the connected ECUs, hardware resources available at remote locations can be fully automated.

Historical driver behavior and drive style are crucial inputs in addition to V2X connectivity data to predict future events as well as fuel consumption of the vehicle on a trip. A trip is a combination of different maneuvers a driver executes to navigate a route and interact with his/her environment including traffic, geography, topography, and weather. This study leverages big data analytics on real-world customer driving data to develop analytical modeling methodologies and algorithms to extract maneuver-based driving characteristics and generate a corresponding maneuver distribution. The distributions are further segmented by additional categories such as customer group and type of vehicle. These maneuver distributions are used to build an aggressivity distribution database which will serve as the parameter basis for further analysis with traffic simulation models.

The work presented here is part of the research done in the field of voice biometrics. This paper helps to understand the state-of-the-art in speaker recognition technology potentially capable of solving challenges related to speaker identification (to identify a speaker among multiple speakers) and speaker verification/authentication (to recognize the current speaking person at a pre-defined access level and authenticate accordingly). The research was focused on performing an unbiased evaluation of two individual voice biometric services. The level of accuracy in identifying and authenticating individuals using these services provides an insight into the current state of technology and the state of what other dual authentication methods could be used to achieve a desired True Acceptance Rate (TAR) and False Acceptance Rates (FAR).

Connected vehicle data unlock compelling solutions for vehicle owners and fleet managers. In selecting machine learning algorithms for use in predicting a connected vehicle signal value, time series dependency is critical to understand. With little to no time series dependency, conventional machine learning models may be used with a feature set that has few or no lag variables. If there is a lot of time series dependency including long-term dependencies, deep learning architectures like variants of recurrent neural networks (RNN) may be a better approach. Further, at any time step, RNN features may be specified to use some number of past time steps to predict the latest value. This paper seeks to identify time series dependency of connected vehicle signals, and selection of the number of time steps to look back in the features set to minimize error.

Electrochemical models of lithium ion batteries are today a standard tool in the automotive industry for activities related to the computer-aided engineering design, analysis, and optimization of energy storage systems for electrified vehicles. One of the challenges in the development or use of such models is the need of detailed information on the cell and electrode geometry or properties of the electrode and electrolyte materials, which are typically unavailable or difficult to retrieve by end-users. This forces engineers to resort to “hand-tuning” of many physical and geometrical parameters, using standard cell-level characterization tests. This paper proposes a method to provide information and data on individual electrode performance that can be used to simplify the calibration process for electrochemical models.