Safe Vehicle Research Articles

Exploring the Role of Transformer Models in Vision, Multi-Modal, and Orbital Data for Autonomous Driving

Abstract. Autonomous driving aims to reduce human error in driving, improve traffic efficiency, and provide a more comfortable driving experience. The integration of computer vision, advanced sensors, and machine learning has been pivotal in this advancement. The introduction of Transformer models has particularly revolutionized the field by offering a novel approach to processing data through attention mechanisms, which is crucial for tasks involving complex relationships between data elements. The paper categorizes research into three main approaches based on input data types: camera-based perception, multi-modal data fusion, and orbital data integration. As autonomous driving technology progresses towards higher levels of autonomy, with L2+ systems becoming standard, challenges remain in accurately interpreting complex environments, handling edge cases, and navigating legal and regulatory landscapes. The paper concludes that while Artificial Intelligence (AI) and deep learning advancements have brought autonomous driving closer to full realization, further research is necessary to address current limitations and ensure safe and reliable autonomous vehicle operation.

Exploring the Role of Transformer Models in Vision, Multi-Modal, and Orbital Data for Autonomous Driving

Abstract. Autonomous driving aims to reduce human error in driving, improve traffic efficiency, and provide a more comfortable driving experience. The integration of computer vision, advanced sensors, and machine learning has been pivotal in this advancement. The introduction of Transformer models has particularly revolutionized the field by offering a novel approach to processing data through attention mechanisms, which is crucial for tasks involving complex relationships between data elements. The paper categorizes research into three main approaches based on input data types: camera-based perception, multi-modal data fusion, and orbital data integration. As autonomous driving technology progresses towards higher levels of autonomy, with L2+ systems becoming standard, challenges remain in accurately interpreting complex environments, handling edge cases, and navigating legal and regulatory landscapes. The paper concludes that while Artificial Intelligence (AI) and deep learning advancements have brought autonomous driving closer to full realization, further research is necessary to address current limitations and ensure safe and reliable autonomous vehicle operation.

Biologics-based technologies for highly efficient and targeted RNA delivery

The demand for RNA-based therapeutics is increasing globally. However, their use is hampered by the lack of safe and effective delivery vehicles. Here, we developed technologies for highly efficient delivery of RNA cargo into programmable extracellular vesicle-mimetic nanovesicles (EMNV) by fabricating hybrid EMNVs-liposomes (Hyb). Tissue targeting is endowed by highly efficient genetic platforms based on truncated CD63 (ΔCD63) or PTGFRN proteins. For the first time we reveal their efficiency to functionalize EMNVs, resulting in >10-fold enhancement of NP internalization in vitro and >2-fold in vivo. RNA delivery using Hyb demonstrated efficiency of >85% in human and mouse cell lines. Comparative analysis of EMNVs and Hyb lysosome colocalization and stability suggested that Hyb enter the lysosomal compartment and escape over time, whereas EMNVs primarily avoid it. Finally, we used these technologies to generate liver-targeting Hyb loaded with therapeutic siRNA, and demonstrated the robust efficiency of this system in vitro and in vivo. These technologies can be adapted for manufacturing a wide range of next-generation vehicles for highly efficient, safe delivery of RNA into desired organs and tissues for therapeutic and prophylactic applications.

Research on trajectory planning based on optimal control modeling

With the advancement of autonomous driving technology, trajectory planning has emerged as one of the core technologies to achieve safe and efficient vehicle driving. An algorithm for single-vehicle trajectory planning based on the Optimal Control Problem (OCP) is proposed in this paper, which can not only plan a safe and efficient path for autonomous vehicles, but also realize the obstacle avoidance function. Firstly, the OCP is modeled for the cycle trajectory planning problem. Secondly, this paper transforms it into a Nonlinear Programming (NLP) with all discrete methods, and uses the interior point method to solve it. Then, the Y-type intersection is selected as the verification scene, and through a series of simulation experiments, the efficiency and safety of the algorithm in different driving environments are verified. The model and algorithm proposed in this paper are not only applicable to Y-type intersections, but also can be extended to other complex traffic roads, providing a new idea and method for trajectory planning in intelligent transportation systems.

Risk-Aware Lane Change and Trajectory Planning for Connected Autonomous Vehicles Based on a Potential Field Model

To enhance the safety of lane changes for connected autonomous vehicles in an intelligent transportation environment, this study draws from potential field theory to analyze variations in the risks that vehicles face under different traffic conditions. The safe minimum vehicle distance is dynamically adjusted, and a comprehensive vehicle risk potential field model is developed. This model systematically quantifies the risks encountered by connected autonomous vehicles during the driving process, providing a more accurate assessment of safety conditions. Subsequently, vehicle motion is decoupled into lateral and longitudinal components within the Frenet coordinate system, with quintic polynomials employed to generate clusters of potential trajectories. To improve computational efficiency, trajectory evaluation metrics are developed based on vehicle dynamics, incorporating factors such as acceleration, jerk, and curvature. An initial filtering process is applied to these trajectories, yielding a refined set of candidates. These candidate trajectories are further assessed using a minimum safety distance model derived from potential field theory, with optimization focusing on safety, comfort, and efficiency. The algorithm is tested in a three-lane curved simulation environment that includes both constant-speed and variable-speed lane change scenarios. Results show that the collision risk between the target vehicle and surrounding vehicles remains below the minimum safety distance threshold throughout the lane change process, ensuring a high level of safety. Furthermore, across various driving conditions, the target vehicle’s acceleration, jerk, and trajectory curvature remained well within acceptable limits, demonstrating that the proposed lane change trajectory planning algorithm successfully balances safety, comfort, and smoothness, even in complex traffic environments.

A comprehensive review on safe reinforcement learning for autonomous vehicle control in dynamic environments

To operate safely in a dynamic environment, autonomous vehicles must possess the same level of predictive driving abilities as human drivers and must be capable of anticipating the future actions of other dynamic objects in the environment, especially those of neighboring vehicles. The development of safe autonomous vehicles (AVs) poses a challenging task as it requires algorithms that can make real-time decisions in unpredictable circumstances. Reinforcement learning (RL) presents a promising approach for AV control, as it utilizes trial and error to enable optimal decision-making. However, traditional RL algorithms are unsuitable for safety-critical applications, as they may explore unsafe actions, potentially resulting in accidents. Safe reinforcement learning (SRL) algorithms have been developed to address this issue, prioritizing safe and reliable decisions. These algorithms incorporate constraints to prevent unsafe actions or utilize techniques to estimate action risk and avoid actions deemed excessively risky. Despite computational challenges, SRL holds significant promise for AV control, and is likely to play a crucial role in developing safe and reliable systems. SRL methods are critical for the general adoption of autonomous vehicles by guaranteeing their safety and reliability. These algorithms utilize methods like uncertainty and risk estimation along with penalty functions, to avoid excessively risky actions and have the potential to significantly reduce accidents and build public trust in autonomous driving. However, there are challenges that need to be addressed, such as the dynamic nature of real-world traffic, high computational costs, and the diversity of road design; and these varying conditions make the designing, testing, and validating of SRL algorithms difficult. Despite these challenges, SRL presents a promising solution, through integrating new sensing technologies and machine learning techniques, to develop safe, efficient, and environmentally friendly transportation systems.

Delivering large genes using adeno-associated virus and the CRE-lox DNA recombination system.

Adeno-associated virus (AAV) is a safe and efficient gene delivery vehicle for gene therapies. However, its relatively small packaging capacity limits its use as a gene transfer vector. Here, we describe a strategy to deliver large genes that exceed the AAV's packaging capacity using up to four AAV vectors and the CRE-lox DNA recombination system. We devised novel lox sites by combining non-compatible and reaction equilibrium-modifying lox site variants. These lox sites facilitate sequence-specific and near-unidirectional recombination of AAV vector genomes, enabling efficient reconstitution of up to 16kb of therapeutic genes in a pre-determined configuration. Using this strategy, we have developed AAV gene therapy vectors to deliver IFT140, PCDH15, CEP290, and CDH23 and demonstrate efficient production of full-length proteins in cultured mammalian cells and mouse retinas. Notably, AAV-IFT140 gene therapy vectors ameliorated retinal degeneration and preserved visual functions in an IFT140-associated retinitis pigmentosa mouse model. The CRE-lox approach described here provides a simple, flexible, and effective platform for generating AAV gene therapy vectors beyond AAV's packaging capacity.

Biopharmaceutical modulatory effects of newly engineered bile acid-Tyloxapol nanogels for attenuation of cytotoxicity in auditory cells

This study presents novel thermoresponsive nanogels composed of chenodeoxycholic acid and Tyloxapol for potential inner ear drug delivery. The nanogels exhibit non-Newtonian, shear-thinning fluid behaviour and rapid gelation at body temperature. Biocompatibility studies were conducted on auditory and macrophage cell lines. Nanogels had a minimal impact on cellular viability, glycolysis, and mitochondrial respiration of auditory cells after 24 h of exposure. However, mitochondrial function analysis in macrophages revealed a significant decrease in oxidative phosphorylation and coupling efficiency after nanogel exposure, accompanied by increased proton leakage. Despite these metabolic disruptions, glycolysis remained unaffected. The Poloxamer matrix appears to be responsible for these effects, independent of the presence of bile acids or Tyloxapol. This study highlights the potential of bile acid-enriched nanogels in drug delivery, particularly their biocompatibility and injectability while acknowledging challenges related to mitochondrial dysfunction in immune cells. These findings suggest that thermoresponsive bile acid and Tyloxapol nanogels have the potential to be safe and effective drug delivery vehicles for inner ear applications.

Assessing public trust in road traffic injuries prevention policies in Iran: a cross-sectional study.

Road Traffic Injuries (RTIs) represent the predominant cause of morbidity and mortality globally. Thus, impactful interventions and trustworthy authorities are imperative to mitigate traffic accidents. The present study seeks to assess public trust in traffic accident diminution policies and measures within the country of Iran. This cross-sectional study utilized a questionnaire comprising 31 questions distributed across eight dimensions. The questionnaire's design was informed by a literature review and semi-structured interviews with experts specializing in traffic accidents. The target population consisted of Tabriz city residents, who were selected via convenience sampling. The collected data were then analyzed using SPSS-16 software, employing the T-test, One-way ANOVA, and Pearson correlation coefficient to present the study's results. structural equation modeling (SEM) was conducted using R 4-4-2 programming language. A total of 681 participants completed the questionnaire, yielding a commendable response rate of 88.6%. The majority of participants were male (60.8%), and 86.2% identified as drivers, with 61.8% having experienced driving accidents. The overall score for public trust in RTIs policies was calculated to be 46.9%. The individual scores for specific domains of trust in traffic safety were as follows: safer roads (43.1%), safe vehicle (43.3%), safety laws (48.8%), safe user (51.1%), safe technology (51.5%), road safety management (46.8%). All the latent variables except education are significantly effective on public trust. The study findings indicate relatively low levels of public trust in RTIs policies among the Iranian population. These results emphasize the need for targeted interventions to enhance public confidence in specific aspects of traffic safety. Policymakers can use these insights to implement effective measures, thus contributing to the RTIs and the promotion of road safety in the country.

Metal-free ring-opening polymerization for the synthesis of biocompatible star-shaped block copolymers with controllable architecture

Star-shaped block copolymers (SBCs) have sparked interest as efficient cargo carriers due to their high loading capacity, decreased burst effects through sustained release, and maintained stability in dilute aqueous solution. Despite these advantages, the practical usage of SBCs is hindered by their challenging synthesis processes that often utilize metal-based catalysts at high temperatures. Herein we report the tailored synthesis of 3-, 4-, and 6-arm polycaprolactone-b-poly(ethylene glycol), PCL-b-PEG, SBCs using diphenyl phosphate as a friendlier, more sustainable non-metallic catalyst. Nuclear magnetic resonance (NMR) analysis confirms the molecular architecture of SBCs and gel permeation chromatography (GPC) is used to elucidate trends in molar mass when the number of arms within the SBCs is tuned, while dynamic light scattering (DLS) and small-angle X-ray scattering (SAXS) studies provide insights into aggregation behavior. Critical aggregation concentration (CAC) values, as measured by fluorescence spectroscopy, demonstrated that the 4-arm and 6-arm SBCs have greater stability than the 3-arm SBC. Biocompatibility assessment indicated minimal cytotoxicity of the nanoaggregates, even at high concentration, making these PCL-b-PEG SBCs potentially safe and sustainable vehicles for biomedical release applications.

ENSURING ECOLOGICAL AND FIRE SAFETY DURING THE OPERATION OF MOTOR VEHICLES WITH PISTON PNEUMATIC AND CRYO ENGINES WHICH USING THE NITROGEN AS WORKING BODY

In the study, the purpose of which is to improve the mathematical apparatus based on the modified thermodynamic theory of disturbances for describing the thermophysical characteristics of nitrogen in the gaseous and liquid states as the working fluid of fire-safe and environmentally safe motor vehicles with piston pneumatic or cryoengines, a mathematical model was suggested to describe for describing the thermodynamic properties of gaseous and liquid of nitrogen, which is intended for the development of a new type of fire-safe and environmentally efficient vehicle with a piston pneumo- or cryoengine and its application is illustrated. The object of the study is the thermophysical characteristics of nitrogen in a liquid or gaseous aggregate state, as a working fluid of fire-safe and eco-safe motor vehicles with a piston pneumatic or cryoengine. The subject of the research is a mathematical apparatus based on the modified perturbation theory for describing the thermophysical characteristics of nitrogen in gaseous and liquid states. The scientific novelty of the research results is that a mathematical apparatus based on a modified thermodynamic theory of disturbances has been improved for a comprehensive description of all thermophysical characteristics of nitrogen, which is in both liquid and gaseous aggregate state, in terms of reducing calculation time and reducing the error of obtaining thermophysical characteristics compared to reference and experimental data. The practical significance of the research results is that the improved mathematical apparatus is marketable for providing accurate information to the composition of the initial data set in studies on the use of nitrogen in various aggregate states as the working fluid of fire-safe and eco-safe motor vehicles with a piston pneumatic or cryoengine.

(Invited) Chiral Graphene Quantum Dot Enhanced Drug Loading into Exosomes

As nanoscale extracellular vesicles secreted by cells, exosomes have enormous potential as safe and effective vehicles to deliver drugs into lesion locations. Despite promising advances with exosome=based drug delivery systems, there are still challenges to drug loading into exosomes, which hinder the clinical applications of exosomes. Herein, we report an exogenous drug-agnostic chiral graphene quantum dots (GQDs) exosome-loading platform, based on chirality matching with the exosomal lipid bilayer. Both hydrophobic and hydrophilic chemical and biological drugs can be functionalized or adsorbed onto GQDs by π–π stacking and van der Waals interactions. By tuning the ligands and GQD size to optimize its chirality, we demonstrate drug loading efficiency of 66.3% and 64.1% for doxorubicin and siRNA, which is significantly higher than other reported sEV loading techniques. These drug/siRNA loaded exosomes can be applied to cancer treatment and immunotheraoy, which showed significant higher efficacy due to the efficient encapsulation of drugs or genetic materials.As nanoscale extracellular vesicles secreted by cells, exosomes have enormous potential as safe and effective vehicles to deliver drugs into lesion locations. Despite promising advances with exosome=based drug delivery systems, there are still challenges to drug loading into exosomes, which hinder the clinical applications of exosomes. Herein, we report an exogenous drug-agnostic chiral graphene quantum dots (GQDs) exosome-loading platform, based on chirality matching with the exosomal lipid bilayer. Both hydrophobic and hydrophilic chemical and biological drugs can be functionalized or adsorbed onto GQDs by π–π stacking and van der Waals interactions. By tuning the ligands and GQD size to optimize its chirality, we demonstrate drug loading efficiency of 66.3% and 64.1% for doxorubicin and siRNA, which is significantly higher than other reported sEV loading techniques. These drug/siRNA loaded exosomes can be applied to cancer treatment and immunotheraoy, which showed significant higher efficacy due to the efficient encapsulation of drugs or genetic materials.

A novel model for real-time risk evaluation of vehicle–pedestrian interactions at intersections

Safety decisions for vehicles at an intersection rely on real-time, objective and continuous assessment of risks in vehicle–pedestrian interactions. Existing surrogate safety models, constrained by ideal assumptions of constant current speed and reliant on interaction points, often misjudge risks, and show inefficiency, inaccuracy and discontinuity. This work proposes a novel model for evaluation of those risks in vehicle–pedestrian interactions at intersections, which abstracts the pedestrian distribution density around a vehicle into a generalized model of driver-pedestrian interaction preferences. The introduction of two conceptions: ’driving risk index’ and ’driving risk gradient,’ facilitates the delineation of driving spaces for identifying safety–critical events. By means of the trajectory data from three intersections, model parameters are calibrated and a multidimensional vehicle–pedestrian interaction risk (VPIR) model is proposed to adapt the complex and dynamic characteristics of vehicle–pedestrian interactions at intersections. Commonly used surrogate safety models, such as Time to Collision (TTC), are selected as benchmark models. Results show that the proposed model overcomes the limitations of the existing interaction-point-based models, and offers a ideal assessment of driving risks at intersections. Finally, the model is illustrated with a case study that assesses the risks in vehicle–pedestrian interactions in varied scenarios and the case study indicates that the VPIR model works well in evaluating vehicle–pedestrian interaction risks. This work can facilitate humanoid learning in the autonomous driving domain, and achieve an ideal evaluation of vehicle–pedestrian interaction risks for safe and efficient vehicle navigation through an intersection.

Enhanced SOC estimation of lithium ion batteries with RealTime data using machine learning algorithms

Accurately estimating Battery State of Charge (SOC) is essential for safe and optimal electric vehicle operation. This paper presents a comparative assessment of multiple machine learning regression algorithms including Support Vector Machine, Neural Network, Ensemble Method, and Gaussian Process Regression for modelling the complex relationship between real-time driving data and battery SOC. The models are trained and tested on extensive field data collected from diverse drivers across varying conditions. Statistical performance metrics evaluate the SOC prediction accuracy on the test set. Gaussian process regression demonstrates superior precision surpassing the other techniques with the lowest errors. Case studies analyse model competence in mimicking actual battery charge/discharge characteristics responding to changing drivers, temperatures, and drive cycles. The research provides a reliable data-driven framework leveraging advanced analytics for precise real-time SOC monitoring to enhance battery management.

Convex Approach to Real-Time Multiphase Trajectory Optimization for Urban Air Mobility

Electric vertical takeoff and landing technology has emerged as a promising solution to alleviate ground transportation congestion. However, the limited capacity of onboard batteries enforces hard time constraints for vehicle operations in a complex urban environment. Therefore, a constrained, real-time trajectory optimization method is necessary to enable safe and energy-efficient vehicle flight. Unfortunately, most existing trajectory optimization methods suffer from low computational efficiency, unpredictable convergence processes, and strong dependence on a good initial trajectory. In this paper, we employ a successive convex programming approach to solve the multiphase minimum-energy-cost electric vertical takeoff and landing vehicle trajectory optimization problem for urban air mobility applications. The main contribution is the development and validation of two novel convexification methods that find approximate optimal solutions to the multiphase nonlinear trajectory optimization problem in real time by solving a sequence of convex subproblems. Specifically, the first method transforms the original nonlinear problem into a sequence of second-order cone programming problems through a convenient change of variables and lossless convexification, while the second approach achieves a similar goal without the need for control convexification. The resulting convex subproblems can be solved reliably in real time by advanced interior point methods. The proposed methods are demonstrated through numerical simulations of two different urban air mobility scenarios and compared with the solution obtained from GPOPS-II, a state-of-the-art general-purpose optimal control solver. Our proposed sequential convex programming methods can obtain near-optimal solutions with faster computational speed than GPOPS-II.

Selection of Li-Ion Batteries Used On Electric Vehicles

The lithium-ion battery is one such energy source. Li-Ion batteries are widely used in EVs for three reasons: they are more energy efficient than conventional batteries, have a longer lifespan, and charge batteries more quickly. The following is a succinct description of the Li-Ion battery's operating principle: The battery's active component is the electrochemical cell, consisting of a cathode and an anode, which are separated and connected by an electrolyte. The primary function of the electrolyte is to facilitate the movement of ions. When the battery is discharged, lithium ions migrate from the anode through the electrolyte to the cathode, while the accompanying electron powers an electrical device. During the charging process, electrons travel from the cathode to the anode through the separator, while current flows from the anode to the cathode. The performance of Li-Ion batteries is primarily influenced by the materials used for the cathode and anode. Early in the nineteenth century, electric vehicles (EVs) using batteries were designed and manufactured. However, compared to fossil fuel cars, whose ingenuity and development outshone EVs, it lagged behind in producing high power. At the moment, internal combustion engines powered by fossil fuels are being phased out in favour of electrical motors for traction because of climate change. Zero CO2 emissions, development on materials has advanced significantly thanks to nanotechnology, and battery development is no exception. On the creation of nanostructured materials for electrodes used in li-ion batteries, several research studies have been described. To provide safe, sturdy, and trustworthy automotive vehicle systems, reliable systems are essential. The selection of Li-ion batteries for use in electric vehicles (EVs) holds significant research significance. Li-ion batteries are currently the preferred choice for EVs due to their high energy density, long cycle life, and relatively low self-discharge rate. However, there are several key research areas that contribute to the importance of battery selection: In this Research we will be using Weighted product method and Weighted sum method. Alternate Parameters taken as LCOB, LNMCOB, LMOB, LFPB, LTOB. Evaluation Parameters taken as Reliability, Safety, Specific power, Specific energy density, Price. The study that was done is briefly summarised in the following lines, along with a prediction of potential future actions. It discusses the evaluation process that was put together before going through the findings related to the utilisation of Li-NMC and Life P batteries in electric cars.

Examining drivers’ distracted driving behaviour using the extended norm activation model

A driving distraction is any external or internal stimulus that diverts a driver’s attention away from the primary task of safe vehicle operation, potentially increasing the risk of accidents. Drivers often get involved in distractions such as phone calls, text messaging, interacting with in-vehicle technology, roadside advertisements, roadside accident sites, and chatting with passengers, etc. In previous research, the antecedents of distraction behaviour have primarily been investigated through the theory of planned behaviour (TPB). Refraining from distracted driving can be considered a pro-social behaviour and can be modelled using the norm activation model (NAM). Therefore, this study aimed to evaluate the potential of pro-social behaviour in explaining drivers’ distracted driving behaviour using the NAM theory. In addition to the conventional factors of the NAM model – awareness of consequences (AC), ascription of responsibility (AR), and personal norms (PN) – this study presents a novel element that examines how technology can mitigate distractions. The data was collected in Lahore city through a questionnaire survey. The results indicate that AC and AR have positive associations with PN, however, PN has also been found to have a positive association with the distraction behaviour. This study provides insights into the utilisation of the NAM theory to evaluate the potential of altruistic behaviour in preventing individuals from engaging in distracted driving.

Multi-Task Visual Perception for Object Detection and Semantic Segmentation in Intelligent Driving

With the rapid development of intelligent driving vehicles, multi-task visual perception based on deep learning emerges as a key technological pathway toward safe vehicle navigation in real traffic scenarios. However, due to the high-precision and high-efficiency requirements of intelligent driving vehicles in practical driving environments, multi-task visual perception remains a challenging task. Existing methods typically adopt effective multi-task learning networks to concurrently handle multiple tasks. Despite the fact that they obtain remarkable achievements, better performance can be achieved through tackling existing problems like underutilized high-resolution features and underexploited non-local contextual dependencies. In this work, we propose YOLOPv3, an efficient anchor-based multi-task visual perception network capable of handling traffic object detection, drivable area segmentation, and lane detection simultaneously. Compared to prior works, we make essential improvements. On the one hand, we propose architecture enhancements that can utilize multi-scale high-resolution features and non-local contextual dependencies for improving network performance. On the other hand, we propose optimization improvements aiming at enhancing network training, enabling our YOLOPv3 to achieve optimal performance via straightforward end-to-end training. The experimental results on the BDD100K dataset demonstrate that YOLOPv3 sets a new state of the art (SOTA): 96.9% recall and 84.3% mAP50 in traffic object detection, 93.2% mIoU in drivable area segmentation, and 88.3% accuracy and 28.0% IoU in lane detection. In addition, YOLOPv3 maintains competitive inference speed against the lightweight YOLOP. Thus, YOLOPv3 stands as a robust solution for handling multi-task visual perception problems. The code and trained models have been released on GitHub.

Optimal Parking Space Selection and Vehicle Driving Decision for Autonomous Parking System Based on Multi-Attribute Decision

Autonomous parking systems (APSs) can help drivers complete the task of finding a parking space and the parking operation, which improves driving comfort. Current research on APSs focus on the perception, localization, planning, and control modules, while few pay attention to the decision modules. This paper proposes a method for optimal parking space selection and vehicle driving decisions. In terms of selecting the optimal parking space, a multi-attribute decision method is designed considering the type of parking space, walking distance, and other factors. In terms of vehicle driving decisions, we first predict the behavior and trajectory of the target vehicle in a specific scenario, and then use a combination of rule-based and learning-based decision methods for safe and comfortable vehicle driving behavior decisions. Simulation results show that the proposed methods can find the optimal parking space according to the parking lot map and improve the efficiency and smoothness of vehicle driving while ensuring driving safety.

Development and testing of immersive virtual reality environment for safe unmanned aerial vehicle usage in construction scenarios

Robotics and autonomous systems are increasingly being used on construction sites to complete tasks and complement human effort. One such autonomous system is unmanned aerial vehicles (UAV), although their wrong use could be potentially hazardous to humans and the environment. Given its relative novelty in construction, there remains a dearth of knowledge about safety risks associated with their use as well as knowledge on skills and protocols for their safe operation. With construction being one of the most dangerous and accident-prone sectors, there is even more impetus to develop UAV safety competence. In this study, we explore the application of immersive virtual reality (VR) use for purposes of safety training pertaining to the use of UAVs in construction scenarios. Whereas, previous studies, focussed on effectiveness of VR as a tool for training individuals on UAV operations and safety risks in general, this study evaluates its effectiveness across distinctive themes of safety in order to compare its effectiveness in different domains of UAV safety. The study adopts design science for development of bespoke UAV safety training tool in VR using Building Information Modelling (BIM) and game engine-driven virtual prototyping. The immersive training tool was then applied in an experiment of participants (n = 100) with (n = 50) constituting a control group using more traditional training methods and a series of pre/post-test assessments. The findings confirmed the relative superiority of VR training over traditional methods, improving performance and retention by up to 22 % on average. The immersive VR was found to be most effective on participants retention of knowledge in general flight preparation, air traffic documentation, visibility management and warning signals. Further usability tests were performed on all participants (n = 100) revealing general presence and positive attitudes towards VR based UAV safety training when compared to traditional training. From a practice perspective this research proofs the effectiveness of a VR approach as more cost effective and safe approach to UAV training for both academia and practice with added advantage of being more realistic simulation of the UAV risk scenarios.