Search Results

Prediction of vehicle velocity is important since it can realize improvements in the fuel economy/energy efficiency, drivability, and safety. Velocity prediction has been addressed in many publications. Several references considered deterministic and stochastic approaches such as Markov chain, autoregressive models, and artificial neural networks. There are numerous new sensor and signal technologies like vehicle-to-vehicle and vehicle-to-infrastructure communication that can be used to obtain inclusive datasets. Using these inclusive datasets of sensors in deep neural networks, high accuracy velocity predictions can be achieved. This research builds upon previous findings that Long Short-Term Memory (LSTM) deep neural networks provide low error velocity prediction. We developed an LSTM deep neural network that uses different groups of datasets collected in Fort Collins, Colorado.

Accurate perception of the driving environment and a highly accurate position of the vehicle are paramount to safe Autonomous Vehicle (AV) operation. AVs gather data about the environment using various sensors. For a robust perception and localization system, incoming data from multiple sensors is usually fused together using advanced computational algorithms, which historically requires a high-compute load. To reduce AV compute load and its negative effects on vehicle energy efficiency, we propose a new infrastructure information source (IIS) to provide environmental data to the AV. The new energy–efficient IIS, chip–enabled raised pavement markers are mounted along road lane lines and are able to communicate a unique identifier and their global navigation satellite system position to the AV. This new IIS is incorporated into an energy efficient sensor fusion strategy that combines its information with that from traditional sensor.

While machine learning in autonomous vehicles development has increased significantly in the past few years, the use of reinforcement learning (RL) methods has only recently been applied. Convolutional Neural Networks (CNNs) became common for their powerful object detection and identification and even provided end-to-end control of an autonomous vehicle. However, one of the requirements of a CNN is a large amount of labeled data to inform and train the neural network. While data is becoming more accessible, these networks are still sensitive to the format and collection environment which makes the use of others’ data more difficult. In contrast, RL develops solutions in a simulation environment through trial and error without labeled data. Our research expands upon previous research in RL and Proximal Policy Optimization (PPO) and the application of these algorithms to 1/18th scale cars by expanding the application of this control strategy to a full-sized passenger vehicle.

Developing a complete understanding of the structure and behavior of the near-wall region (NWR) in reciprocating, internal combustion (IC) engines and of its interaction with the core flow is needed to support the implementation of advanced combustion and engine operation strategies, as well as predictive computational models. The NWR in IC engines is fundamentally different from the canonical steady-state turbulent boundary layers (BL), whose structure, similarity and dynamics have been thoroughly documented in the technical literature. Motivated by this need, this paper presents results from the analysis of two-component velocity data measured with particle image velocimetry near the head of a single-cylinder, optical engine. The interaction between the NWR and the core flow was quantified via statistical moments and two-point velocity correlations, determined at multiple distances from the wall and piston positions.

Today’s Advanced Driver Assistance Systems (ADAS) predominantly utilize cameras to increase driver and passenger safety. Computer vision, as the enabler of this technology, extracts two key environmental features: the drivable region and surrounding objects (e.g., vehicles, pedestrians, bicycles). Lane lines are the most common characteristic extracted for drivable region detection, which is the core perception task enabling ADAS features such as lane departure warnings, lane-keeping assistance, and lane-centering. However, when subject to adverse weather conditions (e.g., occluded lane lines) the lane line detection algorithms are no longer operational. This prevents the ADAS feature from providing the benefit of increased safety to the driver. The performance of one of the leading computer vision system providers was tested in conditions of variable snow coverage and lane line occlusion during the 2020-2021 winter in Kalamazoo, Michigan.

This paper takes a realistic approach to develop a techno-economic analysis for fixed-route autonomous shuttles. To develop a model for analysis, the current state of technology was used to approximate three timelines for achieving SAE level 5 capabilities: progressive, realistic, and conservative. Within these timelines, there are four different increments for advancements in the technology laid out as follows: SAE Level 0 - human driver, SAE Level 4 - in-vehicle safety operator, SAE Level 4 - remote safety operator, and SAE Level 5 - no safety operator. These increments in the changes of the technology were chosen based on the trends in the industry. Various shuttle models were used based on different rider quantities and drive-train requirements (electric vs gas) in this analysis. This allows for further understanding of how these deployment plans will vary the cost for shuttles operating in high, mid, and low ridership demand environments.

Prediction based optimal energy management systems are a topic of high interest in the automotive industry as an effective, low-cost option for improving vehicle fuel efficiency. With the continuing development of connected and autonomous vehicle (CAV) technology there are many data streams which may be leveraged by transportation stakeholders. The Suite of CAVs-derived data streams includes advanced driver-assistance (ADAS) derived information about surrounding vehicles, vehicle-to-vehicle (V2V) communications for real time and historical data, and vehicle-to-infrastructure (V2I) communications. The suite of CAVs-derived data streams have been demonstrated to enable improvements in system-level safety, emissions and fuel economy.

Fatigue damage sensing and monitoring of any structure is a prerequisite for reliable and effective structural health diagnosis. The designed sensor has alternate slots and strips with different strain magnification factor with respect to the nominal strain at its location. The strips experience the strains which closely resemble the actual strain distribution in the critical area of the component. One of the major advantages of this sensor is that it can be placed at any convenient location, still experiencing the same fatigue damage as a critical location. It can be used on various structures from ground civilian and military vehicles to steel bridges. This can predict the remaining useful life of a component or the number of miles (for any automobile) left for the component before it needed replacement. This paper mainly describes the design aspects of this sensor following analytical and finite element analysis (FEA) approaches.

Pressure vessels are being widely employed worldwide as a means to carry, store or receive fluids. The pressure differential is dangerous and many fatal accidents have occurred in the history of their development and operation. Therefore, it is imperative to understand the behavioral effect of cylindrical pressure vessel with torispherical endclosure subjected to an internal pressure. In this paper, two dimensional static stress analyses are performed using the finite element method for different vessel thicknesses in order to understand the stresses and deflections in the vessel walls due to internal pressure. From the analysis, it is observed that the stress variation over the section of the geometry and thickness of the vessel play an important role in withstanding the applied internal pressure.

Estimated engine torque is an important parameter used by automotive systems for automated transmission and clutch control. Heavy-duty engine and transmission manufacturers widely use SAE J -1939 based ECU torque calculation based on mass air/fuel flow steady state maps created during calibration of the engine for this purpose. As an alternative, to enhance the accuracy of this important control variable, a virtual flywheel torque sensor (VFTS) was developed. It measures the engine torque based on the harmonics of the instantaneous flywheel speed signal. Initial dynamometer testing showed the VFTS estimated torque values exhibited a maximum inaccuracy of 12% of the actual measured torque over the range of conditions tested. In this paper we report the results of on road truck testing of the VFTS. A loaded heavy truck with a gross vehicle weight rating of 80,000 pounds was used.

Autonomous vehicles rely upon accurate information about their surrounding environment to perform safe operational planning. The environment sense and perception system normally produces camera image data and LiDAR point cloud data that are processed and then fused to obtain a better perception of the environment than is possible from either alone. The accuracy of the fused data depends upon knowledge of the position of each sensor on the ego vehicle. Vehicle damage, improper sensor installation, sensor mount deformation, mount movement excited by vehicle motion, and/or other situations can result in an unexpected position of the sensor. This error adds uncertainty into the sensor measurement fusion that is normally not accounted for. LiDAR translational offset and angular orientation misalignment errors are investigated for correction.

Lane detection plays a critical role in autonomous vehicles for safe and reliable navigation. Lane detection is traditionally accomplished using a camera sensor and computer vision processing. The downside of this traditional technique is that it can be computationally intensive when high quality images at a fast frame rate are used and has reliability issues from occlusion such as, glare, shadows, active road construction, and more. This study addresses these issues by exploring alternative methods for lane detection in specific scenarios caused from road construction-induced lane shift and sun glare. Specifically, a U-Net, a convolutional network used for image segmentation, camera-based lane detection method is compared with a radar-based approach using a new type of sensor previously unused in the autonomous vehicle space: radar retro-reflectors.

Practical applications of recently developed sensor fusion algorithms perform poorly in the real world due to a lack of proper evaluation during development. Existing evaluation metrics do not properly address a wide variety of testing scenarios. This issue can be addressed using proactive performance measurements such as the tools of resilience engineering theory rather than reactive performance measurements such as root mean square error. Resilience engineering is an established discipline for evaluating proactive performance on complex socio-technical systems which has been underutilized for automated vehicle development and evaluation. In this study, we use resilience engineering metrics to assess the performance of a sensor fusion algorithm for vehicle localization. A Kalman Filter is used to fuse GPS, IMU and LiDAR data for vehicle localization in the CARLA simulator.

Contemporary ADS and ADAS localization technology utilizes real-time perception sensors such as visible light cameras, radar sensors, and lidar sensors, greatly improving transportation safety in sufficiently clear environmental conditions. However, when lane lines are completely occluded, the reliability of on-board automated perception systems breaks down, and vehicle control must be returned to the human driver. This limits the operational design domain of automated vehicles significantly, as occlusion can be caused by shadows, leaves, or snow, which all occur in many regions. High-definition map data, which contains a high level of detail about road features, is an alternative source of the required lane line information. This study details a novel method where high-definition map data are processed to locate fully occluded lane lines, allowing for automated path planning in scenarios where it would otherwise be impossible.

Automotive companies are constantly looking to increase the fuel efficiency, shift quality, passenger comfort, and to reduce wear and tear on the components. Most of these aspects depend on the accuracy of torque used for transmission control, which determines the required operational gear position at a given speed and road conditions. Currently, SAE J-1939 CAN bus torque estimation relies on steady state maps that are generated during the calibration of the engine for different speeds and loads. In this paper we report the development of a Virtual Flywheel Torque Sensor (VFTS) useful for real time torque measurement based on an engine speed harmonics analysis. The VFTS uses a signal from the flywheel speed sensor to estimate the flywheel angular acceleration, which and provides a proportional torque value which corresponds to torque at the flywheel.

Standard operation of autonomous vehicles on public roads results in significant exposure to high levels of risk. There is a significant need to develop metrics that evaluate safety of an automated system without reliance on the rate of vehicle accidents and fatalities compared to the number of miles driven; a proactive rather than a reactive metric is needed. Resilience engineering is a new paradigm for safety management that focuses on evaluating complex systems and their interaction with the environment. This paper presents the overall methodology of resilience engineering and the resilience assessment grid (RAG) as an evaluation tool to measure autonomous systems' resilience. This assessment tool was used to evaluate the ability to respond to the system. A Pure Pursuit controller was developed and utilized as the path tracking control algorithm, and the Carla simulator was used to implement the algorithm and develop the testing environment for this methodology.

Operation of an autonomous vehicle (AV) carries risk if it acts on inaccurate information about itself or the environment. The perception system is responsible for interpreting the world and providing the results to the path planning and other decision systems. The perception system performance is a result of the operating state of the sensors, e.g. is a sensor in fault or being adversely affected by the weather or environmental conditions, and approach to sensor measurement interpretation. We propose a trailing horizon switched system observer that minimizes the difference between reference tracking values developed from sensor fusion performed at an upper level and the values from a potentially faulty sensor based upon the convex combination of different sensor observation model outputs; the sensor observations models are associated with different sensor operating errors.

Development of autonomous vehicle technology is expensive and perhaps more complicated than initially thought, as evidenced by the recent rollback of anticipated delivery dates from companies such as Tesla, Waymo, GM, and more. One of the most effective techniques to reduce research and development costs and speed up implementation is rigorous analysis through simulation. In this paper, we present multiple autonomous vehicle perception and control strategies that are rigorously investigated in the user friendly, free, and open-source simulation environment, CARLA. Overall, we successfully formulated potential solutions to the autonomous navigation problem and assessed their advantages and disadvantages in simulation at no cost. First, a lane finding method utilizing polynomial fitting and machine learning is proposed. Then, the waypoint navigation strategy is described, along with route planning. Object detection is then implemented using pre-trained convolutional neural networks.

A path tracking controller is essential for an autonomous vehicle to navigate a complex environment while avoiding obstacles. Many research studies have proposed new controller designs and strategies. However, it is often unclear which control strategy is the most suitable for a specific Autonomous / ADAS user application. This study proposes a benchmark workflow by comparing different control observer models and their control strategies integration for an autonomous parking lot sweeper in a complex and dense environment at low-speed utilizing model-in-the-loop simulation. The systematic procedure consists of the following steps: (1) vehicle observer model validation (2) control strategy development (3) model-in-the-loop simulation benchmark for specific user scenarios. The kinematic and dynamic vehicle models were used to validate the truck’s behavior using physical data.

Transportation vehicle and network system efficiency can be defined in two ways: 1) reduction of travel times across all the vehicles in the system, and 2) reduction in total energy consumed by all the vehicles in the system. The mechanisms to realize these efficiencies are treated as independent (i.e., vehicle and network domains) and, when combined, they have not been adequately studied to date. This research aims to integrate previously developed and published research on Predictive Optimal Energy Management Strategies (POEMS) and Intelligent Traffic Systems (ITS), to address the need for quantifying improvement in system efficiency resulting from simultaneous vehicle and network optimization. POEMS and ITS are partially independent methods which do not require each other to function but whose individual effectiveness may be affected by the presence of the other. In order to evaluate the system level efficiency improvements, the Mobility Energy Productivity (MEP) metric is used.