Hasil untuk "Industrial safety. Industrial accident prevention"

Mohamad Xazaquan Mansor Ali, Mohammad Lui Juhari, K. Arifin et al.

Confined spaces present serious work risks because of restricted access, inadequate ventilation, and exposure to poisonous or oxygen-deficient environments, which frequently result in injuries and fatalities. Incidents still happen because of incomplete safety interventions, poor training, and gaps in hazard knowledge, even in the face of legislative frameworks like ISO 45001 and OSHA 29 CFR 1910.146. In order to find trends, important contributors, and theme clusters, this study conducts a thorough bibliometric analysis of international research on confined space safety and health using the Scopus database and VOSviewer software. The analysis covered 147 peer-reviewed publications between 1940 and February 2025. Four dominant research clusters were identified: (1) Human and Occupational Health (HOH), focusing on demographic risks and injuries in sectors such as agriculture; (2) Safety Management and Regulations (SMR), emphasizing safety frameworks and compliance; (3) Occupational Exposure and Workplace Safety (OEWS), addressing exposure risks and preventive measures; and (4) Hazards, Risk Assessment, and Prevention (HRAP), covering toxic gases, ventilation strategies, and industrial hygiene practices. The study also highlights an increasing trend in using real-time gas monitoring, IoT-enabled safety technologies, and AI-based predictive tools for confined space hazard mitigation. However, there are still significant research gaps in fields not well covered in the literature today, such as biological risks, ergonomic design, and psychological stressors. In order to improve confined space safety, the study emphasises the necessity of multidisciplinary approaches that integrate engineering, behavioural science, and regulatory insights. Researchers, legislators, and business professionals looking to create evidence-based safety procedures, promote technological advancement, and lower workplace accidents might use the findings as a starting point. This bibliometric study offers prospects for improving the safety of confined spaces and helps to understand the research evolution better.

Marcos Kalinowski, Lucas Romao, Ariane Rodrigues et al.

Lean R&D has been used at PUC-Rio to foster industry-academia collaboration in innovation projects across multiple sectors. This industrial experience paper describes recent experiences and evaluation results from applying Lean R&D in partnership with Petrobras in the oil and gas sector and Americanas in retail. The findings highlight Lean R&D's effectiveness in transforming ideas into meaningful business outcomes. Based on responses from 57 participants - including team members, managers, and sponsors - the assessment indicates that stakeholders find the structured phases of Lean R&D well-suited to innovation projects and endorse the approach. Although acknowledging that successful collaboration relies on various factors, this industrial experience positions Lean R&D as a promising framework for industry-academia projects focused on achieving rapid, impactful results for industry partners.

Azmine Toushik Wasi, Enjamamul Haque Eram, Sabrina Afroz Mitu et al.

Industry 5.0 marks a new phase in industrial evolution, emphasizing human-centricity, sustainability, and resilience through the integration of advanced technologies. Within this evolving landscape, Generative AI (GenAI) and autonomous systems are not only transforming industrial processes but also emerging as pivotal geopolitical instruments. We examine strategic implications of GenAI in Industry 5.0, arguing that these technologies have become national assets central to sovereignty, access, and global influence. As countries compete for AI supremacy, growing disparities in talent, computational infrastructure, and data access are reshaping global power hierarchies and accelerating the fragmentation of the digital economy. The human-centric ethos of Industry 5.0, anchored in collaboration between humans and intelligent systems, increasingly conflicts with the autonomy and opacity of GenAI, raising urgent governance challenges related to meaningful human control, dual-use risks, and accountability. We analyze how these dynamics influence defense strategies, industrial competitiveness, and supply chain resilience, including the geopolitical weaponization of export controls and the rise of data sovereignty. Our contribution synthesizes technological, economic, and ethical perspectives to propose a comprehensive framework for navigating the intersection of GenAI and geopolitics. We call for governance models that balance national autonomy with international coordination while safeguarding human-centric values in an increasingly AI-driven world.

Afrah Gueriani, Hamza Kheddar, Ahmed Cherif Mazari et al.

The increased Internet of Medical Things IoMT and the Industrial Internet of Things IIoT interconnectivity has introduced complex cybersecurity challenges, exposing sensitive data, patient safety, and industrial operations to advanced cyber threats. To mitigate these risks, this paper introduces a novel transformer-based intrusion detection system IDS, termed BiGAT-ID a hybrid model that combines bidirectional gated recurrent units BiGRU, long short-term memory LSTM networks, and multi-head attention MHA. The proposed architecture is designed to effectively capture bidirectional temporal dependencies, model sequential patterns, and enhance contextual feature representation. Extensive experiments on two benchmark datasets, CICIoMT2024 medical IoT and EdgeIIoTset industrial IoT demonstrate the model's cross-domain robustness, achieving detection accuracies of 99.13 percent and 99.34 percent, respectively. Additionally, the model exhibits exceptional runtime efficiency, with inference times as low as 0.0002 seconds per instance in IoMT and 0.0001 seconds in IIoT scenarios. Coupled with a low false positive rate, BiGAT-ID proves to be a reliable and efficient IDS for deployment in real-world heterogeneous IoT environments

Kévin Ducharlet, Liwen Zhang, Sara Maqrot et al.

Industrial timetabling is a critical task for decision-makers across various sectors to ensure efficient system operation. In real-world settings, it remains challenging because unexpected events often disrupt execution. When such events arise, effective rescheduling and collaboration between humans and machines becomes essential. This paper presents a recommendation system-based framework for handling rescheduling challenges, built on Timefold, a powerful AI-driven planning engine. Our experimental study evaluates nine instances inspired by a realworld preventive maintenance use case, aiming to identify the heuristic that best balances solution quality and computing time to support near-optimal decisionmaking when rescheduling is required due to unexpected events during operational days. Finally, we illustrate the complete process of our recommendation system through a simple use case.

Peter Leisztner, Ferenc Farago, Gyula Szabo

The European Union Council’s zero vision aims to eliminate workplace fatalities, while Industry 4.0 presents new challenges for occupational safety. Despite HR professionals assessing managers’ and employees’ competencies, no system currently exists to evaluate the competencies of workers’ representatives in occupational safety and health (OSH). It is crucial to establish the necessary competencies for these representatives to avoid their selection based on personal bias, ambition, or coercion. The main objective of the study is to identify the competencies and their components required for workers’ representatives in the field of occupational safety and health by following the steps of the DACUM method with the assistance of OSH professionals. First, tasks were identified through semi-structured interviews conducted with eight occupational safety experts. In the second step, a focus group consisting of 34 OSH professionals (2 invited guests and 32 volunteers) determined the competencies and their components necessary to perform those tasks. Finally, the results were validated through an online questionnaire sent to the 32 volunteer participants of the focus group, from which 11 responses (34%) were received. The research categorized the competencies into the following three groups: core competencies (occupational safety and professional knowledge) and distinguishing competencies (personal attributes). Within occupational safety knowledge, 10 components were defined; for professional expertise, 7 components; and for personal attributes, 16 components. Based on the results, it was confirmed that all participants of the tripartite system have an important role in the training and development of workers’ representatives in the field of occupational safety and health. The results indicate that although OSH representation is not yet a priority in Hungary, there is a willingness to collaborate with competent, well-prepared representatives. The study emphasizes the importance of clearly defining and assessing the required competencies.

Sultan Elcin Eroglu, H. Sebnem Duzgun

This study investigates the 2006 Aracoma Alma Mine fire that occurred during an underground coal mining claiming the lives of two miners using a Causal Analysis based on System Theory (CAST). Systemic flaws and serious deficiencies in the mine's control structure were identified as the primary contributing factors to the disaster using the CAST analysis. These results aligned well with the revised Federal Mine Act as well as the MINER Act, which exhibits enhanced emergency response procedures and implemented using safety device like multi-gas detectors and enhanced training programs for mine workers. The Aracoma Alma Mine Fire case was selected in this study for analysis since it has strong resemblance and characteristics with the Soma Mine Disaster (SMD) that occurred in 2014 in Turkey. The aftermath of the accidents revealed that the belt conveyor systems caught fire in both cases and simply the regulations were changed without adopting a robust and sustainable systems approach. In both situations, mining companies considered production ahead of safety due to a prevalent misconception of risk. Analyzing the comparison of CAST assessments of these two cases revealed shared systemic shortcomings which underscores the necessity for a fundamental shift in regulatory strategies that extends beyond mere risk identification and encompass the complexities identified by CAST. In addition to regulatory changes, this study also highlights areas where safety still needs enhancements, such as conducting frequent inspections and improving control structure feedback systems. The comparison of the two mining accidents highlights the need for a comprehensive and multi-layered safety approach that helps build strong safety protocols averting similar incidents in the future.

Krunoslav Borovec

Ekky Imanjaya, Cynthia MF Pangabean

Introduction: As stipulated in the Indonesian Labor Law, every worker is entitled to work safety and health protection, including the film industry. This research focuses on two articles in the Health, Safety, and Environment (HSE) regulations and the Law of Film Year 2009. However, the Indonesian film industry has not officially implemented these laws. There have been several cases of HSE, which caused death or fatal injuries to film workers, without applying the regulations. Other HSE issues include the cases where only a few film producers gave insurance to the film workers, applied proper risk assessment, or provided first aid kits. The paper will overview HSE in the Indonesian film industry by mapping out the problems and potential solutions. Methods: By having in-depth interviews with key persons in the field, such as the workers and film producers, this research aims to map out such issues and answering why and how the laws on work health and safety are not implemented in the Indonesian film industry. Result: This research has resulted in maps of problems and recommendations for policymakers, film workers, and related institutes concerning HSE and the rights of film workers, including of the lack awareness of film workers on HSE and HSE-related curriculum in film education, as well as the need for stronger film associations and union. Conclusion: HSE in the Indonesian film industry must be evaluated to be more effective. Some factors to be reviewed include law enforcement in contracts, health insurance, the collaboration of various parties, HSE-related knowledge in the curriculum in Indonesian film education, and the application of Work Competency Standards (SKKNI) to all film professional associations.

Omar Munaf Tawfeeq, S. Thiruchelvam, Izham Bin Zainal Abidin

Organizations are becoming more aware of the need to ensure a safe working environment for their staff. Technological advancements and industrial growth have enhanced efficiency, however, they present new challenges and risks for employees. Accidents remain a concern despite International Labor Organization (ILO) guidelines, governmental bodies, and industry institutions promoting workplace safety. Therefore, it is crucial to evaluate the determinants of workplace safety performance, particularly in the electrical power industry. This study formulates a theoretical model to assess the predictors of safety practices of managers and staff in the Iraqi electricity sector, extending the safety climate model with four external constructs and a moderating variable. Data were collected from 374 participants using an online questionnaire and the PLS-SEM method for analysis. The factor loadings exceeded the recommended value of 0.7 and internal consistencies were greater than the threshold value of 0.8. The findings showed that the safety performance in the Iraqi electric power sector is influenced by safety communication, safety policy, safety control, prevention planning, and safety commitment. Safety commitment is affected by safety policy, prevention planning, control, and communication, while safety training and safety control were found to be insignificant. Furthermore, safety communication had the most significant effect. The results of this study provide some theoretical and practical implications for employees' safety performance toward their overall safety in the electric power industry.

Xiaomeng Xu, Donghui Li, Guojun Huang et al.

With the burgeoning landscape of new enterprises and business paradigms, industrial and trade enterprises are facing escalating pressure to ensure operational safety. Conventional safety management mechanisms have proven to be inadequate for adapting to the dynamic market demands and intricacies of modern production environments. To improve safety management practices, this study integrates complex network theory to dissect the causal chains underlying accidents in industry and trade enterprises. A network model is established to elucidate the factors contributing to accidents and leverage datasets from safety inspections to construct a repository of latent safety risks. To address deficiencies in extant safety frameworks, a comprehensive safety management evaluation system is formulated, comprising ten primary evaluation indices and 30 secondary metrics. Based on the established frameworks, such as ISO 45001 for occupational health and safety management systems (OHSMS), standardized safety production protocols, and risk hierarchical management and control systems and hidden hazard identification and treatment systems (dual prevention systems), a holistic safety management system (SMS) is synthesized on the basis of system-theoretic accident model and process (STAMP) theory. This systematic approach culminates in a robust framework tailored to modern industrial and trade enterprises, fostering flexibility and efficacy in safety management capabilities. This case analysis underscores the model’s ability to enhance its safety management proficiency, thereby amplifying its relevance in fortifying enterprise operations and fostering sustainable growth. This study represents a pivotal step toward augmenting safety management capacities within the industrial and trade enterprises to safeguard enterprise vitality and advance sustainable business practices.

Muhammad R Tayab, Hamda Al Suwaidi, Maryam Lari et al.

The integration of operations excellence, process safety, and sustainability across both upstream and downstream segments of the oil and gas industry is vital for ensuring long-term resilience and social responsibility. Operations excellence encompasses the efficient and effective management of all aspects of oil and gas operations, from exploration and production to refining and distribution. This includes optimizing processes, maximizing productivity, and minimizing costs while maintaining high standards of safety and environmental stewardship. Process safety involves preventing and mitigating the impact of hazardous events within industrial operations. In the context of oil and gas, this includes measures to prevent oil spills, gas leaks, fires, and other accidents that could harm workers, communities, and the environment. Sustainability in the oil and gas industry involves reducing environmental impact, minimizing carbon emissions, and promoting social responsibility throughout the entire value chain. This includes efforts to improve energy efficiency, reduce water usage, minimize waste generation, and mitigate the ecological footprint of operations. By integrating operations excellence, process safety, and sustainability principles, ADNOC has achieved several benefits: Enhanced Safety: Implementing rigorous safety protocols and operational best practices reduces the risk of accidents and incidents, safeguarding the well-being of workers, communities, and ecosystems. Improved Efficiency: Optimizing processes and reducing waste not only lowers operational costs but also minimizes resource consumption and environmental impact. Regulatory Compliance: Adhering to stringent safety and environmental regulations ensures legal compliance and reduces the risk of penalties or fines. Reputation Management: Demonstrating a commitment to safety and sustainability enhances ADNOC's reputation, fostering trust among stakeholders and investors. Long-Term Viability: By addressing safety and sustainability challenges proactively, companies can future-proof their operations, ensuring resilience in the face of evolving market dynamics and regulatory requirements. Ten historical process safety incident investigations were reviewed, and root causes were mapped with elements process safety & asset integrity framework and elements of Centre for Chemical Process Safety (CCPS) to highlight focus areas to enhance sustainable industrial operations by prevention process safety incidents. Preventing process safety incidents requires a multifaceted strategy that encompasses various stages, from risk identification to continuous improvement. Preventing process safety incidents in the oil and gas industry demands an innovative and holistic approach. This novel strategy encompasses several key components: Advanced Risk Identification and Assessment: Leveraging cutting-edge technologies and methodologies for identifying and assessing risks allows for a more accurate understanding of potential hazards and their implications. Next-Generation Preventive Measures: Incorporating state-of-the-art engineering controls, predictive analytics, and advanced process automation enhances the effectiveness of preventive measures, mitigating risks proactively. Dynamic Barrier Management: Implementing dynamic barrier management systems that adapt to real-time operational conditions ensures the resilience and robustness of safety barriers, minimizing the likelihood of failures. This innovative approach not only addresses traditional aspects such as risk assessment and preventive measures but also integrates cutting-edge technologies and methodologies to enhance the effectiveness and efficiency of process safety strategies in the oil and gas industry.

Olaf Sassnick, Georg Schäfer, Thomas Rosenstatter et al.

Industrial Operational Technology (OT) systems are increasingly targeted by cyber-attacks due to their integration with Information Technology (IT) systems in the Industry 4.0 era. Besides intrusion detection systems, honeypots can effectively detect these attacks. However, creating realistic honeypots for brownfield systems is particularly challenging. This paper introduces a generative model-based honeypot designed to mimic industrial OPC UA communication. Utilizing a Long ShortTerm Memory (LSTM) network, the honeypot learns the characteristics of a highly dynamic mechatronic system from recorded state space trajectories. Our contributions are twofold: first, we present a proof-of concept for a honeypot based on generative machine-learning models, and second, we publish a dataset for a cyclic industrial process. The results demonstrate that a generative model-based honeypot can feasibly replicate a cyclic industrial process via OPC UA communication. In the short-term, the generative model indicates a stable and plausible trajectory generation, while deviations occur over extended periods. The proposed honeypot implementation operates efficiently on constrained hardware, requiring low computational resources. Future work will focus on improving model accuracy, interaction capabilities, and extending the dataset for broader applications.

MohammadNoor Injadat

The continued growth in the deployment of Internet-of-Things (IoT) devices has been fueled by the increased connectivity demand, particularly in industrial environments. However, this has led to an increase in the number of network related attacks due to the increased number of potential attack surfaces. Industrial IoT (IIoT) devices are prone to various network related attacks that can have severe consequences on the manufacturing process as well as on the safety of the workers in the manufacturing plant. One promising solution that has emerged in recent years for attack detection is Machine learning (ML). More specifically, ensemble learning models have shown great promise in improving the performance of the underlying ML models. Accordingly, this paper proposes a framework based on the combined use of Bayesian Optimization-Gaussian Process (BO-GP) with an ensemble tree-based learning model to improve the performance of intrusion and attack detection in IIoT environments. The proposed framework's performance is evaluated using the Windows 10 dataset collected by the Cyber Range and IoT labs at University of New South Wales. Experimental results illustrate the improvement in detection accuracy, precision, and F-score when compared to standard tree and ensemble tree models.

Muhammad Aqeel, Shakiba Sharifi, Marco Cristani et al.

This study introduces the Iterative Refinement Process (IRP), a robust anomaly detection methodology designed for high-stakes industrial quality control. The IRP enhances defect detection accuracy through a cyclic data refinement strategy, iteratively removing misleading data points to improve model performance and robustness. We validate the IRP's effectiveness using two benchmark datasets, Kolektor SDD2 (KSDD2) and MVTec AD, covering a wide range of industrial products and defect types. Our experimental results demonstrate that the IRP consistently outperforms traditional anomaly detection models, particularly in environments with high noise levels. This study highlights the IRP's potential to significantly enhance anomaly detection processes in industrial settings, effectively managing the challenges of sparse and noisy data.

K. Bong, Byung-Jick Kim, Youngmin Choi et al.

Recently, the demand for safety in industrial facilities has been increasing, and many studies are being conducted. However, there is not much research on multi-use facilities. In practice, there is a tendency to focus more on industrial facilities than on multi-use facilities due to economic importance, complex technical issues, compliance with legal regulations and safety standards, and concentration of research resources. This study focuses on safety management in multi-use facilities. Multi-use facilities play an important role in preventing and managing potential risks and accidents in multi-use facilities used by many people. Multi-use facilities, which are public facilities, can cause major civil disasters that result in many casualties when a building collapses or a fire occurs. Therefore, this study analyzed factors affecting safety and health performance centered on multi-use facilities. A survey was conducted on 128 safety management workers working in multi-use facilities, and statistical analysis was conducted on the 128 valid questionnaires using the statistical analysis program IBM SPSS-28. Factors affecting safety and health performance were examined by dividing them into safety and health system, manager role, facility management, and employee safety awareness. Through this study, we examined the factors affecting safety and health performance that can prevent accidents in multi-use facilities used by the general public, and confirmed that the safety and health system, facility management, and employee safety awareness had a significant effect. It is expected that safety accident prevention and efficient management will be achieved in multi-use facilities.

I. V. Shapoval, L. Karimova, G. Tikhonova et al.

Introduction. Every year, about 350 thousand people die in the workplace for reasons related to production all over the world. The significance of this problem dictates the need to analyze fatal occupational injuries in order to further develop a set of measures aimed at preventing it. The study aims to analyze fatal occupational injuries at enterprises of the Republic of Bashkortostan on the basis of personalized data as an information basis for the development and justification of priority areas for the prevention of accidents in the workplace. Materials and methods. For analyzing the indicators of general occupational injuries and fatal injuries, we used the results of the all-Russian monitoring of labor conditions and safety of the Ministry of Labor and Social Protection, Russian Federation, analytical materials of labor conditions and safety of the Ministry of Family, Labor and Social Protection of the Population of the Republic of Bashkortostan for the period 2017-2020. To assess the completeness of the accounting of occupational injuries in Russia and the Republic of Bashkortostan, the researchers used the ILO methodology "On assessing the reliability of statistics of accidents at work in countries with imperfect accounting". Based on the materials of 177 Acts on the investigation of fatal accidents (Form 4) provided by the State Labor Inspectorate in the Republic of Bashkortostan for the period 2017-2020, we have studied the circumstances and causes of the death of workers at work, their professional status and age-length characteristics. Results. The analysis of the dynamics of occupational injuries for 2017-2020 in the Russian Federation and the Republic of Bashkortostan showed a decrease in both total occupational injuries and fatal injuries, with a decrease in the frequency of worker deaths occurring at a faster pace. This provided an increase in the ratio of the total number of injuries to the number of fatal injuries, indicating an increase in the level of safety at enterprises and the quality of accounting for minor injuries. However, in 2020 the level of fatal industrial injuries in the Republic exceeded the same indicator in Russia by 25%. An in-depth analysis of fatal injuries based on accident investigation materials in the Republic of Bashkortostan for 2017-2020 showed that most often workers died in construction (0.77 per 1000 workers) and mining enterprises (0.75%) as a result of such types of accidents as traffic accidents, falling from a height, exposure to moving objects, flying rotating objects, parts, machines, etc. Specialists also observed a high level of fatal injuries in agriculture (0.58%), transport and storage enterprises (0.41%), water supply, sewerage (0.38%), etc. The main causes of fatal injuries were unsatisfactory organization of work (34.7%) and violation of traffic rules (29.2%). At the victim’s workplaces the researchers have identified a significant number of violations of labor protection requirements: the absence of special assessment of working conditions (SAWC), briefings and training on labor protection, violations of the work and rest regime, labor and industrial discipline, non-issuance of personal protective equipment (PPE) and the absence of mandatory preliminary and periodic medical examinations. Male workers were most often died (97.6%); in the profession of "driver"; at the age of 30-39 years. There is a very high proportion of victims with work experience of up to one year (44.6%). Almost 75% of the victims had less than 5 years of work experience. This indicates unsatisfactory training in occupational safety of newly hired workers, regardless of their age and previous experience at other enterprises. Conclusion. The in-depth analysis of fatal injuries at enterprises of the Republic of Bashkortostan showed the need to develop a set of targeted occupational safety measures aimed at reducing the level of occupational injuries, taking into account the most traumatic types of economic activity, the most frequent types of accidents and causes of accidents. In addition, special attention should be paid to the training of safe methods and techniques for performing the work of low-skilled workers, as well as the organization of checking the knowledge of traffic rules among drivers of vehicles.

E. Sugak

In accordance with the decisions of the Government of the Russian Federation, the occupational safety management system is being reformed in the country. The new model of this system, built on the basis of occupational risk management techniques, in particular, using the Deming — Shewhart cycle, makes it possible to qualitatively improve the prevention of occupational injuries. To methodologically support the transition to a new model of labor protection, the main regulatory documents were prepared and approved, and additions were made to the Labor Code of the Russian Federation. However, the introduction of preventive mechanisms to improve working conditions is hampered by the problem of taking into account injuries of light and moderate severity. The established long-term practice of ignoring a significant number of industrial hazards conflicts with the requirements of mandatory implementation of procedures for managing occupational risks, which affects the efficiency of the activities carried out. In particular, in the European Union countries 10–40 times more accidents are registered than in Russia, but at the same time the injury severity rate there is 10 times lower. The International Labor Organization, using a special indicator on fatal injury statistics, suggested that the real number of people injured at work in Russia could range from 600 thousand to 1.2 million people in year. Using the example of Heinrich injury pyramid, the methods for recognizing industrial hazards that are in an obvious and hidden state are considered. Evidence-based statistics on the positive impact of the number of identified hazards on workplace safety are also provided. Taking into account the existing practice of risk accounting, it is proposed to use the injury severity coefficient as the main indicator of production safety. At the same time, one should strive not to reduce the injury frequency rate, but, on the contrary, to increase it, which will indicate an improvement in the recording of injuries.

Olav Lamberts, Konrad Wolsing, Eric Wagner et al.

Industrial systems are increasingly threatened by cyberattacks with potentially disastrous consequences. To counter such attacks, industrial intrusion detection systems strive to timely uncover even the most sophisticated breaches. Due to its criticality for society, this fast-growing field attracts researchers from diverse backgrounds, resulting in 130 new detection approaches in 2021 alone. This huge momentum facilitates the exploration of diverse promising paths but likewise risks fragmenting the research landscape and burying promising progress. Consequently, it needs sound and comprehensible evaluations to mitigate this risk and catalyze efforts into sustainable scientific progress with real-world applicability. In this paper, we therefore systematically analyze the evaluation methodologies of this field to understand the current state of industrial intrusion detection research. Our analysis of 609 publications shows that the rapid growth of this research field has positive and negative consequences. While we observe an increased use of public datasets, publications still only evaluate 1.3 datasets on average, and frequently used benchmarking metrics are ambiguous. At the same time, the adoption of newly developed benchmarking metrics sees little advancement. Finally, our systematic analysis enables us to provide actionable recommendations for all actors involved and thus bring the entire research field forward.

Noah Goodall

Between June 2018 and December 2022, Tesla released quarterly safety reports citing average miles between crashes for Tesla vehicles. Prior to March 2021, crash rates were categorized as 1) with their SAE Level 2 automated driving system Autopilot engaged, 2) without Autopilot but with active safety features such as automatic emergency braking, and 3) without Autopilot and without active safety features. In January 2022, Tesla revised past reports to reflect their new categories of with and without Autopilot engaged, in addition to making small adjustments based on recently discovered double counting of reports and excluding previously recorded crashes that did not meet their thresholds of airbag or active safety restraint activation. The revisions are heavily biased towards no-active-safety-features$\unicode{x2014}$a surprising result given prior research showing that drivers predominantly keep most active safety features enabled. As Tesla's safety reports represent the only national source of Level 2 advanced driver assistance system crash rates, clarification of their methods is essential for researchers and regulators. This note describes the changes and considers possible explanations for the discrepancies.