Vehicle Sensor Data Research Articles

A least-squares identification method for vehicle cornering stiffness identification from common vehicle sensor data

In this study, we present a robust framework for vehicle tire cornering stiffness identification, leveraging commonly available lateral acceleration and yaw rate measurements. In contrast to most of the state-of-the-art approaches, our methodology does not rely on (augmented) state estimation, but on batch least-squares optimisation of larger time windows. This batch-approach has as a major benefit that many of the observability issues which are encountered in estimation-based methods are completely eliminated. By immediately exploiting the measurement equations, rather than comparing dynamic simulation responses, we obtain a very efficient and sparse formulation which enables online processing. The methodology was validated on (openly available) datasets from two different vehicles – a Ferrari 250LM and Range Rover Evoque, delivering consistent and accurate results for both vehicles. This work shows that batch optimisation can be a very promising alternative for the more common state-estimation approaches for extracting reliable vehicle cornering stiffnesses in various driving scenarios with a much more straightforward tuning.

Online LiDAR-camera extrinsic parameters self-checking and recalibration

Abstract During production, smart cars are equipped with calibrated LiDARs and cameras. However, due to the vibrations that inevitably occur during driving, the sensors’ extrinsic parameters may change slightly over time. It is a significant challenge to ensure the ongoing security of these systems throughout the car’s lifetime. To address this issue, we propose a self-checking and recalibration algorithm that can continuously detect the sensor data of intelligent vehicles. If the sensor’s miscalibration is detected, the data can be repaired promptly to ensure the vehicle’s reliability. Our self-checking algorithm extracts features from the point cloud and image and performs pixel-wise comparisons. To improve feature quality, we utilize the patch-wise transformer to enhance local information exchange, which also benefits the subsequent extrinsic recalibration. To facilitate the study, we generate two customized datasets from the KITTI dataset and the Waymo Open Dataset. The experiments conducted on these datasets demonstrate the effectiveness of our proposed method in accurately calibrating the LiDAR and camera systems throughout the car’s lifetime. This study is the first to highlight the importance of continually checking the calibrated extrinsic parameters for autonomous driving. Our findings contribute to the broader goal of improving safety and reliability in autonomous driving systems. The dataset and code are available at https://github.com/OpenCalib/LiDAR2camera_self-check.

A systematic review for the fatigue driving behavior recognition method

Fatigue driving is one of the primary causative factors of road accidents. It is of great significance to discern, identify and warn drivers in time for traffic safety and reduce traffic accidents. In this paper, a systematic review for the fatigue driving behavior recognition method is developed to analyze its research status and development trends. Firstly, the data information and its application scenarios related to fatigue driving is detailed. Three driving behavior recognition methods based on different types of signal data are summarized and analyzed, and this signal data can be divided into physiological signal characteristics, visual signal characteristics, vehicle sensor data characteristics and multi-data information fusion. By summarizing and comparing the recognition effect of existing fatigue driving recognition methods, combined with deep learning technology, the paper concludes the fatigue driving behavior recognition method based on single data source has some shortcomings such as low accuracy and easy to be affected by external factors, but the recognition method based on multi-feature information fusion can achieve a exhilarated recognition result. Finally, some prospects are given to analyze the development trend of fatigue driving behavior recognition in the future.

Analysis of traffic conflicts at big intersection types in urban areas

In the past decades, planners developed intersections for the caroriented city. Above all, they should handle motor vehicle traffic efficiently. Urban areas currently have a high proportion of private cars and, as a result, congestion and high CO2 emissions. At the same time, the importance of and demand for sustainable mobility is increasing. Pedestrians and cyclists are particularly vulnerable as the weakest road users. The goal of transport policy is to achieve “Vision Zero,” i.e., no serious injuries or fatalities in road traffic (Federal Ministry of Digital Affairs and Transport, 2022). Many accidents occur primarily at intersections and junctions. Road traffic authorities in Germany generally use accident data from the accident atlas as the basis for evaluating the road safety of intersections. Planners can improve traffic safety by avoiding critical situations. To this end, an intensive study of the development of conflicts as a precursor to accidents can provide further insights. The research in this project shows that such data on traffic conflicts are not available to road authorities. The project will investigate which and how many conflicts between pedestrian, bicycle, and motor vehicle traffic frequently occur at individual intersection types, such as signalized intersections or traffic circles. The initial goal of the project is to develop a survey concept. In the future, planners should be able to use this to quickly identify risks to pedestrian and bicycle traffic at the types of intersections studied. With the help of vehicle sensor data (e.g. from emergency brake assistants), further findings on the frequency and course of conflicts are to be derived. In addition, a methodology would be developed that will enable traffic planners and municipalities to carry out risk management for intersections in the future in order to improve traffic safety and promote local mobility. From the results of the conflict analysis, we will derive recommendations for transport planning and policy. An action guide will then present the recommendations. The end of the project will be in summer 2024.

A Bayesian extreme value theory modelling framework to assess corridor-wide pedestrian safety using autonomous vehicle sensor data

Pedestrians are a vulnerable road user group, and their crashes are generally spread across the network rather than in a concentrated location. As such, understanding and modelling pedestrian crash risk at a corridor level becomes paramount. Studies on pedestrian crash risks, particularly with the traffic conflict data, are limited to single or multiple but scattered intersections. A lack of proper modelling techniques and the difficulties in capturing pedestrian interaction at the network or corridor level are two main challenges in this regard. With autonomous vehicles trialled on public roads generating massive (and unprecedented) datasets, utilising such rich information for corridor-wide safety analysis is somewhat limited where it appears to be most relevant. This study proposes an extreme value theory modelling framework to estimate corridor-wide pedestrian crash risk using autonomous vehicle sensor/probe data. Two types of models were developed in the Bayesian framework, including the block maxima sampling-based model corresponding to Generalised Extreme Value distribution and the peak over threshold sampling-based model corresponding to Generalised Pareto distribution. The proposed framework was applied to autonomous vehicle data from Argoverse—a Ford Motors subsidiary. This autonomous vehicle fleet of Agro AI (owner of Argoverse dataset) is equipped with two 64 beams synchronised LiDAR sensors, a cluster of seven high-resolution cameras, and a pair of stereo-vison high-resolution cameras to capture surrounding road users’ information within a range of 200 meters. A subset of the Argoverse dataset, focussing on an arterial corridor in Miami, USA, was used to extract pedestrian and vehicle trajectories. From these trajectories, vehicle–pedestrian conflicts were identified and measured using post encroachment time. The non-stationarity of extremes was captured by vehicle volume, pedestrian volume, average vehicle speed, and average pedestrian speed in the extreme value model. Both block maxima and peak over threshold sampling-based models were found to provide a reasonable estimate of historical pedestrian crash frequencies. Notably, the block maxima sampling-based model was more accurate than the peak over threshold sampling-based model based on mean crash estimates and confidence intervals. This study demonstrates the potential of using autonomous vehicle sensor data for network-level safety, enabling an efficient identification of pedestrian crash risk zones in a transport network.

Investigating the influential factors in real-world energy consumption of battery electric vehicles

The transportation industry is undergoing a major shift towards electrification to mitigate carbon emissions and decrease reliance on fossil fuels in response to global climate change and greenhouse gas concerns. Recently, battery electric vehicles (BEVs) have made significant progress and are becoming a more viable option for achieving zero-emission transportation. The aim of this study is to investigate the energy consumption patterns of BEVs operating in real-world driving scenarios encompassing various route conditions. The vehicle sensor data employed in this study was acquired through onboard diagnostic devices that directly gathered raw data and subsequently transmitted it to mobile applications. Various significant factors, such as payload, road slope level, speed range, acceleration, and loads of heating, ventilation, and air conditioning, are considered as variables influencing energy consumption. By utilizing the large amount of data collected, machine learning (ML) techniques were applied to develop a predictive model of energy consumption and identify variables that influence energy consumption. The outcomes of this study hold the potential to offer guidance to transportation policymakers and furnish valuable insights for prospective buyers considering BEVs. Furthermore, the application of ML in the development of a predictive model demonstrated efficacy and exhibits promising potential for wider-ranging applications.

Estimating Energy Consumption of Battery Electric Vehicles Using Vehicle Sensor Data and Machine Learning Approaches

Transport electrification, which entails replacing fossil fuel-powered engines with electric drivetrains through the use of electric vehicles (EVs), has been identified as a potential strategy for reducing emissions in the transportation sector. As the adoption of EVs increases, there is a growing need to understand their performance and characteristics, particularly the factors that influence energy consumption under actual driving conditions. This study sought to investigate the actual energy consumption of commercial battery electric vehicles (BEVs) in Thailand by conducting real-world driving tests under various route conditions, including urban and rural route modes. Data collection was performed through the use of onboard diagnostics and global positioning system devices. The result shows that the average energy consumption of the BEVs in this study was 148.03 Wh/km. Moreover, several machine learning (ML) techniques were utilized to analyze the collected dataset to predict energy consumption and identify the key factors influencing energy consumption. A comprehensive investigation of factor significance was carried out by employing a specific algorithm in conjunction with the SHapley Additive exPlanations (SHAP) approach. This investigation provided insights into the influence of battery current and vehicle speed on the energy consumption of BEVs, particularly in the context of urban route conditions. The results of this study provide valuable insights into the energy consumption of BEVs and the factors affecting it, which can aid in improving energy efficiency and informing policy decisions related to transport electrification.

A Safe Driving Decision-Making Methodology Based on Cascade Imitation Learning Network for Automated Commercial Vehicles

Safe driving decision-making is particularly important for automated commercial vehicles. Small passenger vehicles pay more attention to collision prevention, while commercial vehicles with a longer brake distance and worse roll stability need to consider both anti-collision and anti-rollover. The widely studied driving decision-making methods for small passenger vehicles cannot be simply and directly applied to commercial vehicles. This paper proposes a safe driving decision-making methodology based on a cascade imitation learning network (CILN). The CILN integrates two parts, namely the supervised learning part and the imitation learning part. The first part learns safe driving maneuvers extracted from naturalistic vehicle sensor data. Through sensor data processing, it develops decision-making at a humanoid level such as avoiding jerky driving actions. In the second part, generative adversarial imitation learning is introduced to further learn safe driving decisions under conditions prone to collision and rollover. Finally, both highD dataset and SUMO (Simulation of Urban Mobility) are used to train and verify the performance of the CILN. By comparing the evaluation indicators of TTC, RTTC, DRAC, distance headway, and lateral acceleration, the CILN outperforms the other decision-making algorithms. Experimental results show that the CILN can provide safe driving decision-making and ensure the driving safety of ACVs in dense traffic flow.

XACML for Mobility (XACML4M)-An Access Control Framework for Connected Vehicles.

The automotive industry is experiencing a transformation with the rapid integration of software-based systems inside vehicles, which are complex systems with multiple sensors. The use of vehicle sensor data has enabled vehicles to communicate with other entities in the connected vehicle ecosystem, such as the cloud, road infrastructure, other vehicles, pedestrians, and smart grids, using either cellular or wireless networks. This vehicle data are distributed, private, and vulnerable, which can compromise the safety and security of vehicles and their passengers. It is therefore necessary to design an access control mechanism around the vehicle data's unique attributes and distributed nature. Since connected vehicles operate in a highly dynamic environment, it is important to consider context information such as location, time, and frequency when designing a fine-grained access control mechanism. This leads to our research question: How can Attribute-Based Access Control (ABAC) fulfill connected vehicle requirements of Signal Access Control (SAC), Time-Based Access Control (TBAC), Location-Based Access Control (LBAC), and Frequency-Based Access Control (FBAC)? To address the issue, we propose a data flow model based on Attribute-Based Access Control (ABAC) called eXtensible Access Control Markup Language for Mobility (XACML4M). XACML4M adds additional components to the standard eXtensible Access Control Markup Language (XACML) to satisfy the identified requirements of SAC, TBAC, LBAC, and FBAC in connected vehicles. Specifically, these are: Vehicle Data Environment (VDE) integrated with Policy Enforcement Point (PEP), Time Extensions, GeoLocation Provider, Polling Frequency Provider, and Access Log Service. We implement a prototype based on these four requirements on a Raspberry Pi 4 and present a proof-of-concept for a real-world use case. We then perform a functional evaluation based on the authorization policies to validate the XACML4M data flow model. Finally, we conclude that our proposed XACML4M data flow model can fulfill all four of our identified requirements for connected vehicles.

Conditional Predictive Maintenance of Electric Vehicles from Electrical and Mechanical Faults

Maintenance of the core components of an electric vehicle is very crucial to ensure productivity, longevity, drive quality, and a safe environment. Predictive Maintenance is an approach that uses the operating & faulty condition data to predict future machine conditions and make decisions upon this prediction. The methodology used for predictive maintenance and condition monitoring can be based on machine learning and data analytics. The process of learning starts with the observation of data and using it in future instances for building the model. The primary aim is to allow the computer to learn without the involvement of the intervention of human assistance. A few machine learning methods are supervised learning, semi-supervised learning, and reinforcement learning. The main aim of the presented research is to use the available sensor data of the electric vehicle from various electronic control units and design a predictive model which classifies the various electrical and mechanical faults that occur in an electric vehicle and predicts the types for increasing the reliability of the whole electrical vehicular system. The workflow of the project is defined as fault modelling, generating healthy and fault data, processing the data using time synchronous averaging, identification of the system condition indicators and finally using these condition indicators, an SVM classification prediction model is designed from which the desired results and conclusion are inferred from the simulation studies.

Smart Black Box DVR System in IT-Based Vehicle Emergency Rescue Environment

In this paper, the driving habits and patterns are analyzed by automatically recognizing vehicle accidents and recording event data when dangerous situations occur, applying connected car service through IT technology. In addition, SBDS designs the highly scalable modeling that reproduces videos which are recorded before and after accidents. SBDS provides emergency rescue and traffic information services by receiving IoT sensor data connected to have an embedded system in the vehicle and notifying the accident location and video information to the control center, using wireless network service even when the vehicle is dormant. In addition, it is possible to implement a wireless e-Call service by interlocking ECU information and vehicle sensor data in case an emergency occurs.

Measuring Roadway Lane Widths Using Connected Vehicle Sensor Data.

The United States has over three trillion vehicle miles of travel annually on over four million miles of public roadways, which require regular maintenance. To maintain and improve these facilities, agencies often temporarily close lanes, reconfigure lane geometry, or completely close the road depending on the scope of the construction project. Lane widths of less than 11 feet in construction zones can impact highway capacity and crash rates. Crash data can be used to identify locations where the road geometry could be improved. However, this is a manual process that does not scale well. This paper describes findings for using data from onboard sensors in production vehicles for measuring lane widths. Over 200 miles of roadway on US-52, US-41, and I-65 in Indiana were measured using vehicle sensor data and compared with mobile LiDAR point clouds as ground truth and had a root mean square error of approximately 0.24 feet. The novelty of these results is that vehicle sensors can identify when work zones use lane widths substantially narrower than the 11 foot standard at a network level and can be used to aid in the inspection and verification of construction specification conformity. This information would contribute to the construction inspection performed by agencies in a safer, more efficient way.

Traffic Event Detection from Consumer Vehicle Sensor Data: An Autonomous Vehicle Study

Traffic event detection from vehicles’ on-board sensors can detect the road information and improve traffic management and safety. Current sensor-based traffic event detection is mainly based on probe vehicles, test vehicles, or other designated vehicles, which is costly and cannot be deployed on a large scale. With the fast development of on-board equipment, data collection from consumer vehicle sensors is becoming popular and can cover a large geographic scale with almost no equipment or labor cost. However, there are very few studies of the data features and potential for application. This paper presents a pipeline for employing consumer vehicle sensors to detect roadworks. The unique data features and deficiencies of the consumer vehicle sensor are discussed and summarized. A clustering method is employed to distinguish the roadwork sites. A route builder method is proposed to reconstruct the routes of the roadworks and to extract the corresponding start and end locations. Compared with ground truth, roadworks detection from consumer vehicle sensors can cover up to 86% of the roadworks on freeways and over 40% of the roadworks on non-freeway roads. The average offset errors are 4.7% on freeways, and 24.9% on non-freeway roads. Compared with point-based roadworks detection from “ViaMichelin Traffic information,” the results from consumer vehicle sensors achieved a matching ratio of nearly 90% and were advantageous in extracting the roadworks route information. This study proves the possibility of employing consumer vehicle sensors for route-based traffic event detection and provides insights for countering uneven distribution and trajectory truncation issues related to privacy protection.

Detecting aggressive driving patterns in drivers using vehicle sensor data

Aggressive driving is known to be a cause of vehicle accidents. Individuals with Attention-deficit hyperactivity disorder (ADHD) are prone to more aggressive behavior and that also leads to aggressive driving. To prevent aggressive driving, we strive to first understand aggressive driving and find patterns in this type of driving behavior. In an effort to uncover to identify patterns in aggressive driving, we examine sensor data and video data of trips taken by drivers with ADHD and identify our distinct aggressive driving patterns. Using the sensor data, we extend our findings to all aggressive trips in our dataset and generate a model to detect aggressive driving patterns. By finding the similarity between trips and then using these distances to produce a KNN model, we are able to model our data and classify it into 4 driving patterns. This analysis can better inform us of the type of driving patterns that appear in aggressive driving. Using this analysis, we can also better understand which patterns are produce better precision and recall using this methodology.

Latency Reduction in Vehicular Sensing Applications by Dynamic 5G User Plane Function Allocation with Session Continuity.

Vehicle automation is driving the integration of advanced sensors and new applications that demand high-quality information, such as collaborative sensing for enhanced situational awareness. In this work, we considered a vehicular sensing scenario supported by 5G communications, in which vehicle sensor data need to be sent to edge computing resources with stringent latency constraints. To ensure low latency with the resources available, we propose an optimization framework that deploys User Plane Functions (UPFs) dynamically at the edge to minimize the number of network hops between the vehicles and them. The proposed framework relies on a practical Software-Defined-Networking (SDN)-based mechanism that allows seamless re-assignment of vehicles to UPFs while maintaining session and service continuity. We propose and evaluate different UPF allocation algorithms that reduce communications latency compared to static, random, and centralized deployment baselines. Our results demonstrated that the dynamic allocation of UPFs can support latency-critical applications that would be unfeasible otherwise.

Correction Approaches to Vehicle Sensor Data for Real-world Road Weather Detection

Correction Approaches to Vehicle Sensor Data for Real-world Road Weather Detection

Traffic density estimation using vehicle sensor data

The objective of this study is to analyze the possibility of using vehicle sensors to estimate traffic density in future traffic environments. For the analysis, the simulation experiments used in this study consist of three steps to estimate traffic density based on vehicle sensor data. The first step is to develop software module that can generate error of distance measurement in a similar way to the real road driving. In the second step, this study conducted traffic simulation experiments by combining the true distance value measured in the simulation with the error value calculated by the error generating module of the first step. Finally, we conducted evaluation procedure by comparing the estimation results of traffic density to the true density value. According to the experiment results, the use of front and rear radar showed best performance in that this case maintained low MAPE (Mean Absolute Percentage Error) results even at low probe vehicle ratio as compared to other cases. In case of camera sensor, it can be important alternative at the standpoint of potential future applications, especially under limited traffic environments of high ratio of sensor-equipped vehicle and congested traffic conditions such as LOS (Level of Service) D or E.

Evaluation of LiDAR data processing at the mobile network edge for connected vehicles

5G mobile network technology together with edge computing will create new opportunities for developing novel road safety services in order to better support connected and automated driving in challenging situations. This paper studies the feasibility and benefits of localized mobile network edge applications for supporting vehicles in diverse conditions. We study a particular scenario, where vehicle sensor data processing, required by road safety services, is installed into the mobile network edge in order to extend the electronic horizon of the sensors carried by other vehicles. Specifically, we focus on a LiDAR data-based obstacle warning case where vehicles receive obstacle warnings from the mobile network edge. The proposed solution is based on a generic system architecture. In this paper, we first evaluate different connectivity and computing options associated with such a system using ns-3 simulations. Then, we introduce a proof-of-concept implementation of the LiDAR-based obstacle warning scenario together with first results from an experimental evaluation, conducted both in a real vehicle testbed environment and in a laboratory setting. As a result, we obtain first insights on the feasibility of the overall solution and further enhancements needed.

Ensemble CNN to Detect Drowsy Driving with In-Vehicle Sensor Data.

Drowsy driving is a major threat to the safety of drivers and road traffic. Accurate and reliable drowsy driving detection technology can reduce accidents caused by drowsy driving. In this study, we present a new method to detect drowsy driving with vehicle sensor data obtained from the steering wheel and pedal pressure. From our empirical study, we categorized drowsy driving into long-duration drowsy driving and short-duration drowsy driving. Furthermore, we propose an ensemble network model composed of convolution neural networks that can detect each type of drowsy driving. Each subnetwork is specialized to detect long- or short-duration drowsy driving using a fusion of features, obtained through time series analysis. To efficiently train the proposed network, we propose an imbalanced data-handling method that adjusts the ratio of normal driving data and drowsy driving data in the dataset by partially removing normal driving data. A dataset comprising 198.3 h of in-vehicle sensor data was acquired through a driving simulation that includes a variety of road environments such as urban environments and highways. The performance of the proposed model was evaluated with a dataset. This study achieved the detection of drowsy driving with an accuracy of up to 94.2%.

Conceptual design of a trust model for perceptual sensor data of autonomous vehicles

Abstract This paper presents a conceptual trust model for the perceptual sensor data of autonomous vehicles to improve their reliability and safety of operation. In particular, the model aims to ensure the trustworthiness of the data used in the automated decision-making processes of vehicles. The presented model forms a comprehensive view of factors that affect trust in the perceptual sensor data. These factors, namely environment, operation, security, and technical limitations, further cover a number of related trust parameters used to evaluate trust values. Each parameter is evaluated using Dempster-Shafer theory using a trust-distrust frame. Mass values are based on instantaneous sensor data and sensor behaviour tested under similar circumstances. As a result, this model is viable for use in conjunction with data fusion and decision-making to improve their reliability by enabling the tracking of the development of sensor trustworthiness.