Autonomous Test Research Articles

Reinforcement Learning Informed Evolutionary Search for Autonomous Systems Testing

Evolutionary search (ES)-based techniques are commonly used for testing autonomous robotic systems. However, these approaches often rely on computationally expensive simulator-based models for test scenario evaluation. To improve the computational efficiency of the search-based testing, we propose augmenting the ES with a reinforcement learning (RL) agent trained using surrogate rewards derived from domain knowledge. In our approach, known as RIGAA (Reinforcement learning Informed Genetic Algorithm for Autonomous systems testing), we first train an RL agent to learn useful constraints of the problem and then use it to produce a certain part of the initial population of the search algorithm. By incorporating an RL agent into the search process, we aim to guide the algorithm towards promising regions of the search space from the start, enabling more efficient exploration of the solution space. We evaluate RIGAA on two case studies: maze generation for an autonomous “Ant” robot and road topology generation for an autonomous vehicle lane-keeping assist system. In both case studies, RIGAA reveals more failures of a high level of diversity than the compared baselines. RIGAA also outperforms the state-of-the-art tools for vehicle lane-keeping assist system testing, such as AmbieGen, CRAG, WOGAN, and Frenetic in terms of the number of revealed failures in a two-hour budget.

The Future of Automation Testing: From SDET to Autonomous Testing

The Future of Automation Testing: From SDET to Autonomous Testing

Complexity Evaluation for Urban Intersection Scenarios in Autonomous Driving Tests: Method and Validation

As autonomous driving technology scales up, complex urban intersections pose significant safety challenges. Current testing methods struggle to simulate these complex scenarios at a manageable cost, making simulation testing essential. For effective evaluation, establishing comprehensive and objective complexity metrics is crucial. However, existing complexity evaluation methods often depend on the performance of the primary vehicle and are based on local interaction relationships, which lack a global perspective and objectivity and have yet to be validated by autonomous driving systems. To address this issue, this paper proposes a multidimensional complexity assessment framework that introduces system-level indicators such as vehicle count, interaction density, disorder, and risk. This framework quantifies the complex interactions at intersections from a global perspective, independent of primary vehicle performance. Experimental results demonstrate that the complexity evaluation results are highly consistent with the performance of a high-level autonomous driving system (Apollo). The framework has been successfully applied to test scenario generation on the Apollo platform, achieving twice the scenario generation efficiency of traditional methods, thus showcasing substantial engineering value.

Assessing public a priori acceptance of fully automated vehicles using an extended technology acceptance model and importance-performance analysis

Fully automated vehicles (FAVs) are anticipated to enter the passenger vehicle market soon. Given the uncertainties surrounding user adoption of this emerging technology, research is needed to understand their user acceptance. While most studies on the acceptance of automated vehicles have been conducted in upper-middle-income or high-income developed countries, similar research in middle-income countries is limited. This study aims to evaluate a priori acceptance of FAVs in a middle-income developing country by extending the Technology Acceptance Model (TAM). Trust, subjective norms, perceived safety risk and four decision-making styles (Thoroughness, Hesitancy, Social resistance, and Perfectionism) were included in the extended model. This study aims to evaluate a priori acceptance of FAVs in a middle-income developing country by extending the Technology Acceptance Model (TAM). Trust, subjective norms, perceived safety risk and four decision-making styles (Thoroughness, Hesitancy, Social resistance, and Perfectionism) were included in the extended model. Structural Equation Modeling was applied to confirm model validation by using data from 1026 drivers from different cities in Iran. A multi-group analysis was conducted to assess whether the relationships between model constructs vary across different demographic and background groups. Additionally, an importance-performance analysis was performed to gain a deeper understanding of the factors influencing behavioral intention. The findings of the model highlighted that beyond the original TAM scales, subjective norms were the strongest predictor in explaining drivers' intentions to use FAVs. Initial trust also moderately contributed to explaining user acceptance of FAVs. The results also revealed that among decision-making styles, thoroughness positively affected behavioral intention through trust, while social resistance had an indirect negative effect on intention. The findings showed that effect of the perceived safety risk on behavioral intention through the initial trust was confirmed; however, importance-performance map analysis revealed that a significant improvement was observed in the safety area that could still be obtained. It is suggested that policymakers start promoting the usefulness and ease of use of FAVs through advertisements, social media, public campaigns and autonomous vehicle test ride events to facilitate the adoption of FAVs when available in countries with similar sociocultural contexts.

A városi önvezetőjármű-felkészültség kulcstényezőinek meghatározása strukturált szöveg- és tartalomelemzéssel = Investigating Autonomous Vehicle Readness of Cities: A Structured Text and Content Analysis

The process of urban integration of autonomous vehicles is posing increasing challenges for many cities today. Based on the growing number of autonomous vehicle tests on urban roads, it seems that the urban presence of these vehicles requires only minimal preparation from cities. However, their widespread adoption, integration, and effective and safe operation at the urban level require substantial preparation. This study aims to synthesize the factors found in the literature using software-supported structured text and content analysis. Based on our results, we identified nine main factors and twenty-seven key elements that could be crucial in supporting the future mass adoption of autonomous vehicles in urban areas. With our work, we hope to contribute to a deeper understanding of urban autonomous vehicle readiness and the identification of measures necessary to achieve it.

Risk and Complexity Assessment of Autonomous Vehicle Testing Scenarios

Autonomous vehicles (AVs) must fulfill adequate safety requirements before formal application, and performing an effective functional evaluation to verify vehicle safety requires extensive testing in different scenarios. However, it is crucial to rationalize the application of different scenarios to support different testing needs; thus, one of the current challenges limiting the development of AVs is the critical evaluation of scenarios, i.e., the lack of quantitative criteria for scenario design. This study introduces a method using the Spherical Fuzzy-Analytical Network Process (SF-ANP) to evaluate these scenarios, addressing their inherent risks and complexities. The method involves constructing a five-layer model to decompose scenario elements and using SF-ANP to calculate weights based on element interactions. The study evaluates 700 scenarios from the China In-depth Traffic Safety Study–Traffic Accident (CIMSS-TA) database, incorporating fuzzy factors and element weights. Virtual simulation of vehicles in the scenarios was performed using Baidu Apollo, and the performance of the scenarios was assessed by collecting the vehicle test results. The correlation between the obtained alternative safety indicators and the quantitative values confirms the validity and scientific validity of this approach. This will provide valuable guidance for categorizing audiovisual test scenarios and selecting corresponding scenarios to challenge different levels of vehicle functionality. At the same time, it can be used as a design basis to generate a large number of effective scenarios to accelerate the construction of scenario libraries and promote commercialization of AVs.

EddyBot: A multichannel FPGA-based Eddy Current Testing mobile robot

We introduce EddyBot, a novel integration of a multi-channel FPGA-based Eddy Current Testing (ECT) scanner within a mobile robot designed for autonomous non-destructive testing (NDT). EddyBot distinguishes itself by leveraging advanced signal processing, real-time robotics control, and efficient data synchronization to significantly enhance inspection accuracy and efficiency. Our contributions are (i) the deployment of a Zynq-7020 FPGA for fast data acquisition and processing, (ii) alongside the Robot Operating System (ROS) for precise and real-time control with Kalman filtering (KF) sensor fusion for pose estimation. (iii) We further optimize defect detection and localization through a novel data synchronization approach, correlating the robot’s pose with ECT signals to a mean synchronization delay of only 11.77 ms. This allows for precise localization of defects within 0.18 cm, improving the overall reliability. The robot’s performance was evaluated on aluminum and ferromagnetic specimens, highlighting the system’s capability to detect defects with high accuracy.

Future-Proofing Your Organization: Key Principles for Leading Through the AI Transition

As artificial intelligence (AI) advances rapidly, organizations must develop meaningful long-term strategies to realize opportunities and mitigate risks, as academic research shows AI will significantly impact over 50% of jobs by 2030, requiring workforce planning, while industries like travel agencies have faced disruption from failing to adapt to digital technologies like AI. Effective AI strategies include establishing integrated 3-5 year technology roadmaps that articulate adoption timelines, skill development, and business transformations, identifying core capabilities to augment with AI, such as diagnostics for healthcare, while data and technology foundations are also critical for deploying AI projects at scale and continuous skills training is needed to fill emerging roles and offset potential job losses. The article discusses strategic considerations for industries like healthcare, where strategies focus on augmenting clinicians, standardizing data, and establishing governance, transportation organizations must plan for autonomous vehicle testing and adoption, and reskill mechanics, and financial institutions can develop robo-advisors and lending platforms while modernizing legacy systems. Therefore, organizational leaders need such integrated roadmaps and ongoing governance to leverage AI's productivity benefits while guiding responsible adoption and workforce evolution amid ongoing disruption through proactive strategic planning that will help direct, rather than undergo, technological transformation.

Real‐Time Model Predictive Control of Air‐Conditioners Through IoT—Results From an Experimental Setup in a Tropical Climate

ABSTRACTThere is an increasing demand for air‐conditioners (ACs) to maintain a comfortable indoor environment for all types and sizes of buildings including commercial, industrial, and office spaces. Such ACs are mostly operated by traditional ON–OFF controllers to maintain a temperature setpoint. Extensive control engineering knowledge resulting from experiments on actual buildings is needed before the wide application of an advanced control methods, such as model predictive control (MPC), which are more effective and energy‐efficient than the traditional controllers. Simulation studies on the application of control may provide promising results, but the corresponding experimental validations may prove otherwise. User‐friendly experimental setups to investigate the performance of real‐time advanced control on an actual building and its HVAC system is scarce. This paper details the design, implementation and testing of a user‐friendly, remote and autonomous test bed to acquire measured data through an IoT platform, and to control ACs in real time through MATLAB Thingspeak. Measurement and data acquisition equipment are installed in a two‐zone concrete building in Mauritius. MPC of the indoor air temperature achieved an AC temperature setpoint tracking RMSE which was 0.7°C lower than that achieved by the built‐in ON/OFF AC control. The test bed also revealed that portable air‐conditioners are not very efficient, given that the maximum cooling efficiency achieved in this work was only 60%. It also provided valuable insights based on the experiments carried out, in terms of improvements to sensing and data acquisition. Control engineers can implement such a test bed for the development and application of advanced controllers as per their needs and applications.

High-risk powered two-wheelers scenarios generation for autonomous vehicle testing using WGAN

Objective Autonomous vehicles (AVs) have the potential to revolutionize the future of mobility by significantly improving traffic safety. This study presents a novel method for validating the safety performance of AVs in high-risk scenarios involving powered 2-wheelers (PTWs). By generating high-risk scenarios using in-depth crash data, this study is devoted to addressing the challenge of public road scenarios in testing, which often lack the necessary complexity and risk to effectively evaluate the capabilities of AVs in high-risk situations. Method Our approach employs a Wasserstein generative adversarial network (WGAN) to generate high-risk scenes, particularly focusing on PTW scenarios. By extracting 314 car-to-PTW crashes from the China In-depth Mobility Safety Study–Traffic Accident database, we simulate outcomes using PC-Crash software. The data are divided into scenes at 0.1-s intervals, with WGAN generating numerous high-risk scenes. By using a cumulative distribution function (CDF), we sampled and analyzed the vehicle’s dynamic information to generate complete scenarios applicable to the test. The validation process involves using the SVL Simulator and the Baidu Apollo joint simulation platform to evaluate the AV’s driving behavior and interactions with PTWs. Results This study evaluates model generation results by comparing distributions using Wasserstein distance as an indicator. The generator converges after approximately 200 epochs, with the iterator converging quickly. Subsequently, 10,000 new scenes are then generated. The distribution of several key parameters in the generated scenes can be found to approximate that of the original scenes. After sampling, the usability of generated scenarios is 64.76%. Virtual simulations confirm the effectiveness of the scenario generation method, with a generated scenario crash rate of 16.50% closely reflecting the original rate of 15.0%, showcasing the method’s capacity to produce realistic and hazardous scenarios. Conclusions The experimental results suggest that these scenarios exhibit a level of risk similar to the original crashes and are effective for testing AVs. Consequently, the generated scenarios enhance the diversity of the scenario library and accelerate the overall testing process of AVs.

Exploring Innovative Methods in Maritime Simulation: A Ship Path Planning System Utilizing Virtual Reality and Numerical Simulation

In addressing the high costs, inefficiencies, and limitations of purely digital simulations in maritime trials for unmanned vessel path planning, this paper introduces a ship virtual path planning simulation test system. This system, unbound by temporal and spatial constraints, vividly showcases the navigational performance of vessels. After analyzing the virtual testing requirements for the autonomous navigation performance of unmanned surface vehicles (USVs), we established the overall framework of this system. Data-driven by a numerical simulation platform, the system achieves synchronized operation between physical and virtual platforms and supports interactive path planning simulations between USVs and the virtual testing system. Furthermore, to address the limitations of traditional ship trajectory planning evaluation, this paper develops a global path planning fitness evaluation function that comprehensively considers trajectory safety, navigation distance, and vessel stability, achieving optimal comprehensive routes through the particle swarm optimization algorithm. Test results indicate an average roll reduction of 14.31% in the planned routes, with a slight increase in navigation distance. By integrating two-dimensional curve simulation with three-dimensional visualization, this paper not only overcomes the limitations of purely physical and purely virtual simulations but also enhances the overall credibility and intuitiveness of the simulation. Experimental results validate the system’s effectiveness, providing a novel method for autonomous navigation testing and evaluation of USVs.

CERN for AI: a theoretical framework for autonomous simulation-based artificial intelligence testing and alignment

This paper explores the potential of a multidisciplinary approach to testing and aligning artificial intelligence (AI), specifically focusing on large language models (LLMs). Due to the rapid development and wide application of LLMs, challenges such as ethical alignment, controllability, and predictability of these models emerged as global risks. This study investigates an innovative simulation-based multi-agent system within a virtual reality framework that replicates the real-world environment. The framework is populated by automated 'digital citizens,' simulating complex social structures and interactions to examine and optimize AI. Application of various theories from the fields of sociology, social psychology, computer science, physics, biology, and economics demonstrates the possibility of a more human-aligned and socially responsible AI. The purpose of such a digital environment is to provide a dynamic platform where advanced AI agents can interact and make independent decisions, thereby mimicking realistic scenarios. The actors in this digital city, operated by the LLMs, serve as the primary agents, exhibiting high degrees of autonomy. While this approach shows immense potential, there are notable challenges and limitations, most significantly the unpredictable nature of real-world social dynamics. This research endeavors to contribute to the development and refinement of AI, emphasizing the integration of social, ethical, and theoretical dimensions for future research.

Enhancing road safety through misbehaviour detection in vehicle‐to‐everything systems of Korea

AbstractVehicle‐to‐everything communication systems play a crucial role in enhancing road safety and traffic efficiency through vehicle and roadside infrastructure interactions. To provide robust defences against external threats in secure and trustworthy information exchange, these systems utilise public key infrastructure to authenticate vehicle‐to‐everything participant identities with digital certificates and security credential management systems to administer these certificates and encryption keys. However, even with these defences, vulnerabilities persist, particularly from vehicles with legitimate certificates that may malfunction or be exploited for malicious purposes. To address these issues, this paper introduces a misbehaviour detection (MBD) system, notable for its combined use of local and global MBD algorithms. This system is specifically designed to combat both conventional and novel threats, including slander attacks, in which vehicles with legitimate certificates may be falsely accused, and sophisticated attacks targeting the global MBD system itself. The efficacy of our MBD system was rigorously validated at K‐City, the leading autonomous vehicle technology testing facility in Korea, demonstrating its ability to identify and counter internal misbehaviours precisely.

Initial findings of autonomous radiation mapping robot field test

Radiation mapping using mobile robotic can minimize the risk of operator exposure to radiation while providing detailed radiation maps of targeted areas. Despite its potential advantages, the use of this technology is still limited and requires further investigation. In a previous study, an autonomous radiation mapping robot was designed and developed. This paper presents the initial findings on autonomous radiation mapping field test that focuses on the operation procedures and the challenges faced during the robot deployment. Temporary solutions implemented to overcome the challenges are also discussed. The results will provide insights for future researchers and operators working on mobile robotic technology for radiation mapping.

Evaluation method with digital expert on the criticality of car-following scenarios for autonomous vehicles testing

Evaluation of autonomous vehicles is one of the major challenges before they can be released. Due to the advantages in efficiency, cost, and safety, scenario-based simulation methods have recently received great attention. Even so, as the complexity and uncertainty exist in the real driving environment, the scenarios that autonomous vehicles may encounter are infinite. Therefore, it is necessary to classify simulation scenarios according to their criticality. It contributes to accelerating the evaluation processes. This paper presents a novel criticality evaluation method, based on a proposed Digital Expert, for car-following autonomous driving. The Digital Expert acts as the evaluator to evaluate the criticality of scenarios depending on their driving performance. Driving performance refers to the achieved degree of driving intentions. Firstly, a Digital Expert is established as the evaluator for the criticality of the scenario using the inverse reinforcement learning method. Then, based on the fact that the intention of Digital Expert is to maximize its internal reward function, the reward function is used to evaluate driving performance. Finally, calculating the criticality of the car-following scenario according to the mapping relationship between driving performance and criticality. Using the driving data in the NGSIM data set, this paper generates two groups of simulated car-following scenarios and evaluates the criticalities of the two scenarios. The experimental results show that the proposed criticality evaluation method can reasonably evaluate the criticality of car-following scenarios.

An entropy‐based model for quantifying multi‐dimensional traffic scenario complexity

AbstractQuantifying the complexity of traffic scenarios not only provides an essential foundation for constructing the scenarios used in autonomous vehicle training and testing, but also enhances the robustness of the resulting driving decisions and planning operations. However, currently available quantification methods suffer from inaccuracies and coarse‐granularity in complexity measurements due to issues such as insufficient specificity or indirect quantification. The present work addresses these challenges by proposing a comprehensive entropy‐based model for quantifying traffic scenario complexity across multiple dimensions based on a consideration of the essential components of the traffic environment, including traffic participants, static elements, and dynamic elements. In addition, the limitations of the classical information entropy models applied for assessing traffic scenarios are addressed by calculating magnitude entropy. The proposed entropy‐based model is analyzed in detail according to its application to simulated traffic scenarios. Moreover, the model is applied to real world data within a naturalistic driving dataset. Finally, the effectiveness of the proposed quantification model is illustrated by comparing the complexity results obtained for three typical traffic scenarios with those obtained using an existing multi‐factor complexity quantification method.

Consumers Discernment on Online Shopping

The researcher aims to study Consumers Discernment on Online Shopping, To study the personal factors of consumers that have a relationship with the decision to purchase products. Through the online shopping channel of online consumers in Bangkok ,.to study the marketing mix of consumers that have a relationship with the decision to purchase products online of online consumers in Bangkok., To study consumer perceptions that have a relationship with the decision to purchase products through Online shopping channels of online consumers in Bangkok Population and sample Data were collected from a group of 500 consumers who purchased products through online channels in Bangkok., this research study The researcher has formulated a conceptual framework from the synthesis of concepts, theories, and research related to Consumers Discernment on Online Shopping on the concept ., [2]Kotler, Philip. Marketing management. Pearson Prentice Hall. 2009.), (2023). Consumer Perception towards Online Shopping) Data collection tools/methods The tools used in the research are Questionnaire by collecting data from consumers' perceptions of online shopping. of online consumers in Bangkok by method and using questionnaires as (Questionnaire) through online channels consisting of Consumer perception of online shopping consists of exposure, intention to perceive, and understanding of consumers towards online shopping in Bangkok. showcasing blend comprising of cost, dispersion channels Promoting advancement In terms of clear measurements items, the measurements utilized are recurrence (Recurrence), rate (Rate), cruel (Cruel) and standard deviation (Standard Deviation). Reference measurements examination. The insights utilized are t tests (Autonomous Tests). t-test), One-Way Investigation of Fluctuation (ANOVA), Relationship Examination, and Stepwise Different Relapse Investigation. The results showed The Leadership style and practitioner motivation. Hypothesis 1(H1) Consumer perception of online shopping Consisting of exposure, intention to perceive, and understanding of consumers towards online shopping. In Bangkok Affects the decision to purchase products through the channel Online in Bangkok, DE=0.16**. Accept the research hypothesis. and Hypothesis 2 (H2) The marketing mix consists of price and distribution channels. Marketing promotion Product side of consumers towards online shopping In Bangkok Affects the decision to purchase products through the channel Online in Bangkok DE=0.52*,IE=0.37*Accept.

Deep Learning-Based Inkjet Droplet Detection for Jetting Characterizations and Multijet Synchronization.

Inkjet printing is a powerful direct material writing process. It can be used to deposit microfluidic droplets in designated patterns at submicrometer resolution, which reduces materials usage. Nonetheless, predicting jetting characterizations is not easy because of the intrinsic complexity of the ink-nozzle-air interactions. Thus, inkjet processes are monitored by skilled engineers to ensure process reliability. This is a bottleneck in industry, resulting in high labor costs for multiple nozzles. To address this, we present a deep learning-based method for jetting characterizations. Inkjet printing is recorded by an in situ CCD camera and each droplet is detected by YOLOv5, a 1-stage detector using a convolutional neural network (CNN). The precision, recall, and mean average precision (mAP) at a 0.5 intersection over the union (IoU) threshold of the trained model were 0.86, 0.89, and 0.90, respectively. Each regression result for a detected droplet is accumulated in chronological order for each class of droplet and nozzle. The quantified information includes velocity, diameter, length, and translation, which can be used to synchronize multinozzle jetting and, eventually, the printed patterns. This demonstrates the feasibility of autonomous real-time process testing for large-scale electronics manufacturing, such as the high-resolution patterning of biosensor electrodes and QD display pixels while exploiting big data obtained from jetting characterizations.

Development of an Autonomous Component Testing System with Reliability Improvement Using Computer Vision and Machine Learning

This study evaluated computer vision-based models, including Histogram Analysis, Logistic Regression, Sift-SVM, and Deep learning models, in an autonomous testing system developed for smartphone camera modules. System performance was assessed in a practical factory setting with workers operating the system, and metrics such as processing time, sensitivity, specificity, accuracy, and defect rate were evaluated. Based on the results, the Sift-SVM model demonstrated the greatest potential for enhancing the reliability of the system with a processing time of 0.01578 seconds, a sensitivity of 99.811%, and a reduction in the failure rate to 1888 PPM. The study findings suggest that Sift-SVM has the potential to be practically applied in the industry, thus improving the speed and accuracy of automatic defect detection in manufacturing and reducing the defect rate.

Development of Autonomous Learning Model to Enhance Inflight Safety-Based Competence for Cabin Crew

Currently, the aviation business is a popular service that helps people travel conveniently, quickly, and safely. People in the airline business have become more stressed with the need to learn about inflight safety-based competence to effectively serve the broad range of passengers’ needs. One reason for this is the nature of their jobs, which also affects their learning. Therefore, this research study developed an autonomous learning model to enhance inflight safety-based competence for cabin crew and then studied the effects of using an autonomous learning model to enhance inflight safety-based competence for cabin crew. This study was conducted in two parts using research and development methodology (R&D). The first phase identified problems and needs in learning about inflight safety-based competence, while the second stage examined the results of using the developed learning model. The results revealed that an autonomous learning model consisted of the model’s purpose, identification and management of learning goals, knowledge development of learning strategies, the trainer’s role, practical ideas, and reflection on and evaluation of the learner’s learning. Furthermore, the cabin crew members had improved scores in inflight safety-based competence learning in four areas: safety policy, risk management, safety promotion, and safety assurance, compared before participating in the autonomous learning model testing. The vital learning promotion course for cabin crew dealing with the inflight safety of the passengers. The developed autonomous learning model should enhance the inflight safety-based competence of cabin crew.