Accident Reports Research Articles

A holistic view of maritime navigation accidents and risk indicators: examining IMO reports from 2011 to 2021

This paper investigated the risk indicators in maritime accidents and how they are considered within the reporting of maritime accidents, drawing on ten years of International Maritime Organisation (IMO) (2011–2020) accident reports. It highlighted the lack of consistent findings in studies exploring the role of vessel characteristics in maritime accidents, which often result from different methods, databases, techniques and motivations used by each respective study for gathering and analysing data. Furthermore, as human error continues to be highlighted as the top-cited cause of accidents, this study examined the qualitative content of IMO accident reports in-depth to broaden our understanding of maritime accident risk factors. Using a data-driven approach, statistical (ANOVA) and advanced text-mining techniques (using IRAMUTEQ software) were applied to extract meaning from the semi-structured and unstructured narrative descriptions that constitute most of the national administrations’ investigation reports to the IMO. Building on the text analysis of the IMO accident data, we proposed the Accident Maritime Ecosystem framework, which incorporates individuals, the ship organisation (on board), the internal ship ecosystem (on board and onshore), the external ship ecosystem (external factors) and the global maritime ecosystem (policies and regulations); moreover, it identifies these entities as risk factors in maritime accidents. The findings illustrate how accident reporting is largely human-centric and that as maritime transportation is becoming increasingly complex, there is a need for policy and organisational decision-makers to incorporate a broader scope of actors when considering maritime risk factors, which can be achieved by using the AME framework as a guideline.

Another side of industrial growth in India: Environmental damage from industrial accidents

Environmental damages from man-made hazards like industrial accidents in India have increased in the last decade (2010–2020) which poses a risk to the sustainable industrial development of the nation. This study examines industrial accidents with collateral damage to the environment over a ten-year period from 2010 to 2020. The objective is to understand whether India is normalizing environmental damage from industrial accidents as a side-effect of economic growth. This paper seeks to understand the characteristics and response to industrial accidents with adverse impact on the environment. Using content analysis, we created a unique database of industrial accidents with reported environmental damage using multiple data sources like international disaster database: EM-DAT, Indian media reports, independent disaster reports, Indian government evaluation reports of accidents and company reports. Industrial accidents are categorised by their year of occurrence, the state in India where the accident occurred, accident type, human impact, environmental impact, industry type, ownership of the organisation and the response from society stakeholders. We use the concept of risk society to understand the complex ways the discourse on industrial disasters is shaped and contextualise the findings in the broader development narrative. Our results suggest 560 industrial accidents have occurred with reported environmental damage from 2010 to 2020. Environmental damage was mostly in form of air pollution and water pollution. About 2,500 human lives have been lost and another 8,500 injured during the said period. Based on qualititive and quantitive analysis, we find that Indian states’ industrial profile is related to the number of industrial accidents but find no relationship between the number of accidents and the society’s response. The way forward is to encourage policymakers to ensure that companies, as part of their investment criteria, do the following: spell out appropriate safety measures; take concrete steps to engage in community awareness of the risks involved in the organizations’ operations; and, take responsibility for human and environmental damage in the event of an accident. To reduce the build-up of risk from industrial growth, maintaining and updating a national level database on the environmental damage from industrial accidents is essential. Such a database will also help signal India’s efforts to ensure sustainable manufacturing growth.

Analysis on causes of chemical industry accident from 2015 to 2020 in Chinese mainland: A complex network theory approach

Chemical manufacturing is a long-process industry, where an end product may pass through numerous dangerous and complex steps. In such long chains of coordinated activity, accidents remain common. This study made loss-prevention recommendations for the chemical industry after conducting a review of accident reports and creating a complex network model. A human factor analysis and classification system (HFACS) was used to classify data from 109 investigation reports from the Chinese mainland (2015–2020). Levels Ⅱ and Ⅲ of the HFACS output were fed into a complex network model to generate a map of causes and chains of risk. It was shown that most accidents were directly or indirectly caused by human action, and human factors played a decisive role in occurrence, evolution, and resolution. The model used was visualized in Gephi, and the key cause nodes were identified by their topological characteristics. A modularity algorithm was used to derive the community structures and segment the network map. Crucial nodes in each community were compared with factors for each class in the HFACS model. It was also found that there was a biasing factor in the causal processes of explosive accidents and poisoning and asphyxiation accidents according to the associations classified by modularity. Risk abatement strategies were proposed for the crucial factors.

A data-driven narratives skeleton pattern recognition from accident reports dataset for human-and-organizational-factors analysis

Accidents in the process industry involve several interacting factors, including human and organizational factors (HOFs). A long-standing obstacle to HOFs analysis is lack of data. Accident reports are an essential data source to learn from the past and contain HOFs-related data, but they are usually unstructured text in a not standardized format. Some studies have explored the extraction of information automatically from accident reports based on Natural Language Processing (NLP) techniques. However, they were not dedicated to HOFs. Risk communication is considered an essential pillar in safety and risk science. This research develops a HOFs-focused risk communication framework based on the NLP techniques that can support risk assessment and mitigation. The proposed approach automatically extracts the target groups oriented “Who, When, Where, Why” (4Ws) information from accident reports.This framework was applied to explore the eMARS database. The results show that the “4Ws” skeleton of narratives has appreciated performance in pattern recognition and holistic information analysis. The graphical representation interfaces are designed to display the features of HOFs-related accidents, which can better be communicated to the sharp-end operators and decision-makers.

Collection and classification of influence parameters for safety effectiveness of ADAS

Virtual scenario-based testing has become an acceptable method for evaluating safety effectiveness of advanced driver assistance systems (ADAS). Due to the complexity of the ADAS operating environment, the scenarios that an ADAS could face are almost infinite. Therefore, it is crucial to find critical scenarios to improve the efficiency of testing without compromising credibility. One popular method is to explore the parameterized scenario space using various intelligent search methods. Selecting parameters to parameterize the scenario space is particularly important to achieve good coverage and high efficiency. However, an extensive collection of (relevant) influence parameters is missing, which allows a thorough consideration when selecting parameters regarding specific scenarios. In addition, the general importance definition for individual influence parameters is not provided, regarding the potential influence of their variations on the safety effectiveness of ADAS, which can also be used as a reference while selecting parameters. Combining knowledge from different sources (the published literature, standardized test scenarios, accident analysis, autonomous vehicle disengagement, accident reports, and specific online surveys), this paper has summarized, in total, 94 influence parameters, given the general definitions of importance for 77 influence parameters based on cluster analysis algorithms. The list of influence parameters provides researchers and system developers a comprehensive basis for pre-selecting influence parameters for evaluating the safety effectiveness of ADAS by virtual scenario-based testing and helps check whether certain influence parameters can be a meaningful extension for the evaluation.

AI and IOT Based Road Accident Detection and Reporting System

Abstract: Road accidents are increasing daily as the number of automobiles rises. An annual global death toll of 1.4 million and an injury toll of 50 million are reported by the World Health Organization (WHO). The absence of medical assistance at the scene of the accident or the lengthy response time during the rescue effort are the main causes of mortality. We can reduce delays in a rescue operation that has the potential to save many lives by using a cognitive agent-based collision detection and smart accident alarm and rescue system. To gather and send accident-related data to the cloud or server, the suggested system consists of a force sensor, GPS module, alarm controller, ESP8266 controller, camera, Raspberry Pi, and GSM module. The accident is then verified using cloud- based techniques for deep learning. accident is then verified using cloud-based techniques for deep learning. When the deep learning module notices an accident, it immediately alerts all nearby emergency services, including the hospital, police station, mechanics, etc.

Risk Management in The Packaging Unit at Animal Feed Industry Sidoarjo, Indonesia

Introduction: Animal feed industry Sidoarjo is one of the companies in the animal feed manufacturing industry. The packaging unit is the last unit in the production process that packs the feed ingredients and arranges the finished goods. A location with the most safety hazards based on the work accident reports for 2010-2017 is in the packaging unit as many as 15.78%. The purpose of this research is to analyze the risk management of work safety in the packaging unit of animal feed industry Sidoarjo. Methods: Based on the data collection method, this research was included in observational research with a cross-sectional design. The subjects in this research were workers (37 people) in the animal feed packaging and the Health Safety Environment (HSE) units. The research was conducted at the packaging unit of Animaal Feed Industry, Sidoarjo, Indonesia in November 2019-June 2020. Results: Based on hazard identification, there were 15 hazards with a risk analysis of 53% moderate risk and 47% high risk. The risk evaluation obtained was including 53% tolerable risk and 47% intolerable risk. After controlling efforts, there was a decrease in the risk category. Conclusion: The residual risk obtained after controlling carried out experienced a reduction of moderate risk by 100% and high risk by 71%. The remaining risk was 33% moderate risk and 13% high risk.

Update on Risk Criteria for Crude Oil Tanker Fleet

In recent years, the maritime trade of crude oil has suffered notable perturbations caused by the unbalanced relationship between supply and demand. The COVID-19 pandemic caused a drop in oil consumption in 2019, followed by a reduction in production in 2020. The seaborne transport of oil accounts for approximately 50–60% of all crude oil in world production. The crude oil market is a crucial regulator of the global economy and instabilities in this market have noticeable effects on collective risks. The immediate risks that the society see are the changes in the cost of living, which are followed by political uncertainties. Less visible are the risks that these uncertainties have on shipping companies and the level of management stability they have to maintain in order to keep seagoing safe. This paper presents an update on the overall state of risk management for the crude oil tanker fleet, evidenced by EMSA and other international marine organisations. The previous paper, entitled Safety Assessment of Crude Oil Tankers, which applied the methodology of the Formal Safety Assessment (FSA), was published in 2018 and covered the historical data related to the fleet size, accident reports, amount of oil spilled on sea and the economic value of the crude oil transport business. The particular focus of this paper is on the evolution of the risk acceptance criteria over the years and the difference in the predictions from 2018 to the present day. The effects of the pandemic on crude oil shipping are discussed through the changes in the risks. Three of them are analysed: PLL (potential loss of lives), PLC (potential loss of containment) and PLP (potential loss of property). The representation of the risk applies the F-N curves among the risk acceptance criteria lines observed for different tanker sizes. Among the three risks, the paper exposes the vulnerability of the loss of containment risk, where the strong economic impact of the oil trade outweighs the environmental concerns. In relation to the PLC, the paper proposes the approach of relating the oil spill acceptability with the spill quantity and ship revenue instead of to the cost of cleaning or the cost of environment recovery.

Extraction and analysis of risk factors from Chinese chemical accident reports

Accidents in chemical production usually result in fatal injury, economic loss and negative social impact. Chemical accident reports which record past accident information, contain a large amount of expert knowledge. However, manually finding out the key factors causing accidents needs reading and analyzing of numerous accident reports, which is time-consuming and labor intensive. Herein, in this paper, a semi-automatic method based on natural language process (NLP) technology is developed to construct a knowledge graph of chemical accidents. Firstly, we build a named entity recognition (NER) model using SoftLexicon (simplify the usage of lexicon) + BERT-Transformer-CRF (conditional random field) to automatically extract the accident information and risk factors. The risk factors leading to accident in chemical accident reports are divided into five categories: human, machine, material, management, and environment. Through analysis of the extraction results of different chemical industries and different accident types, corresponding accident prevention suggestions are given. Secondly, based on the definition of classes and hierarchies of information in chemical accident reports, the seven-step method developed at Stanford University is used to construct the ontology-based chemical accident knowledge description model. Finally, the ontology knowledge description model is imported into the graph database Neo4j, and the knowledge graph is constructed to realize the structured storage of chemical accident knowledge. In the case of information extraction from 290 Chinese chemical accident reports, SoftLexicon + BERT-Transformer-CRF shows the best extraction performance among nine experimental models. Demonstrating that the method developed in the current work can be a promising tool in obtaining the factors causing accidents, which contributes to intelligent accident analysis and auxiliary accident prevention.

Spatial-temporal analysis method of ship traffic accidents involving data field: An evidence from risk evolution of ship collision

Collision accident is the commonest type of ship traffic accident. Analysis of the spatial-temporal characteristics during the ship encounter process is helpful to the discovery of the accident evolution mechanism. The spatial-temporal analysis method was developed for collision accidents from maritime accident investigation reports to reveal the evolution characteristics of potential collision risk (PCR) during the two ships' encounters. First, a four-dimensional interpolation algorithm under the Radar window was proposed by analyzing the uncertainty of risk collision from different ship parameters. Second, a data field cognitive model with Gaussian mixture clustering on PCR was constructed according to the Radar information of the ship encounter process. Last but not least, combining different collision avoidance scenarios under the ship domain, the spatial-temporal characteristics of PCR were discussed to reveal the evolution mechanism. Empirical evidence shows that the laws of time are equivalent to space in the four stages of the ship encounter process. Ship's parameters make a difference in risk perception under the Radar window. Different encounter situations have significant impacts on Radar-based collision avoidance.

Involvement of Human and Organizational Factors in Railway Accidents: Application of the Human Factors Analysis and Classification System and Bayesian Network

The occurrence of railway accidents is not always the result of error/violation on the operator’s end; rather, there are many organizational factors involved. This research investigated 97 railway accident reports from the West Zone of Bangladesh Railway between 2015 and 2019 to categorize factors within the institution that contribute to accidents. The accidents were divided into four categories (collision, derailment, locomotive failure, and signal passed at danger [SPAD]) and were applied to the Human Factors Analysis and Classification System (HFACS) framework. The HFACS framework systematically identifies direct and hidden factors that lead to an accident. A total of 225 human and organizational factors were found involved in 97 accidents, of which 106 (47.11%) were organizational-level factors and 119 (52.89%) were front-line/operation-level factors. To understand the association between factors, a Bayesian network was created following the naive Bayes algorithm, keeping “accident type” as the class variable. Analysis revealed that “Collisions” occur because of organizational factors such as the structure of command, planned inappropriate operations, and operation-level factors such as perceptual errors, and personal unreadiness. “Derailment” was influenced by organizational reasons, such as equipment issues, company safety culture, and inadequate supervision, whereas operation-level factors were crew and resource management and skill-based error. “Locomotive failure” was found to be associated with only organizational factors, such as equipment issues and inadequate supervision, and SPAD mostly occurred because of exceptional violations at the operational level. The identified issues, if addressed properly, can help ensure a safer railway industry.

Commercial Aircraft-Assisted Suicide Accident Investigations Re-Visited—Agreeing to Disagree?

Background: The number of aircraft-assisted suicides can only be considered a rough estimate because it is difficult and, at times, impossible to identify all cases of suicide. Methods: Four recent reports of accidents occurring in 1997 in Indonesia, 1999 in Massachusetts in the United States, 2013 in Namibia, and 2015 in France related to commercial aircraft-assisted suicides were analyzed. This analysis relied on data extracted from the accident reports that supported aircraft-assisted suicide from the: (a) cockpit voice recorder (CVR) and flight data recorder (FDR), (b) medical history, (c) psychosocial history, (d) toxicology, (e) autopsy, and (f) any methodology that utilized aviation medicine. There are some limitations in this study. Although all analyzed accident investigations followed ICAO Annex 13 guidelines, there is variability in their accident investigations and reporting. In addition, accident investigation reports represent accidents from 1997 to 2015, and during this time, there has been a change in the way accidents are reported. The nature of this analysis is explorative. The aim was to identify how the various aircraft accident investigators concluded that the accidents were due to suicidal acts. Results: In all four accident reports, FDR data were available. CVR data were also available, except for one accident where CVR data were only partially available. Comprehensive medical and psychosocial histories were available in only one of four of the accident reports. Conclusion: To prevent accidents involving commercial aircraft, it is necessary to identify the causes of these accidents to be able to provide meaningful safety recommendations. A detailed psychological autopsy of pilots can and likely will assist in investigations, as well as generate recommendations that will substantially contribute to mitigating accidents due to pilot suicide. Airborne image recording may be a useful tool to provide additional information about events leading up to a crash and thus assist in accident investigations.

Rome vehicle accidents analysis and city riskiness prediction by Artificial Intelligence

Rome municipality publishes geo-referenced vehicle accidents reports and detailed road traffic counters that, in conjunction with Open Street Map roads details, enable deep analyses of the tragic phenomenon. This study focuses on the full year 2021, during which a total of 62,081 vehicles casualties were registered by Rome traffic authorities. The broad Rome territory, overcoming a bunch of Italian chief towns, the wide and capillary set of arteries and the municipality roads extension greater with respect to many European capitals, enhance the statistical relevance of the analyzed data. Artificial intelligence method takes into account three potential accidents predictors as the level of road traffic, the event timing, and the driving complexity of the different zones. The proposed method has the capability of linking time/traffic/complexity indexes of a specific zone of a city with the accident riskiness, moreover, is applicable to whatever urban complex scenario and takes care of the result format in order to facilitate road safety decision makers in smoothing the tragic sequence of accidents. Further possible improvements directions are discussed.

Analysis and assessment of risks to public safety from unmanned aerial vehicles using fault tree analysis and Bayesian network

Unmanned aerial vehicles (UAVs) have become increasingly prevalent in human production and life, but concerns about their impact on public safety have also grown. To address these concerns, this paper presents a hybrid risk identification model that uses fault tree analysis (FTA) and Bayesian network (BN) to analyze critical risks associated with UAVs. First, by analyzing accident reports and existing literature, the fault tree structure of UAV-related public safety accidents is constructed, and initial risk factors are identified. Next, the fault tree is converted into a BN using conversion rules and the conditional probabilities are determined based on logic gates, parameter learning, and expert opinions. The BN model is then validated using actual UAV accident cases. Finally, diagnostic inference and sensitivity analysis are applied to identify key risk factors that impact public safety from UAVs. The results indicate that the primary risk factors leading to public safety incidents involving UAVs are internal system failures (battery failures, mechanical failures, and flight control system failures), pilot factors (unqualified knowledge and skills, weak safety awareness, violations, and lack of legal awareness), external environmental impacts (obstacles, route planning issues, and unclear airspace division), and UAV supervision issues (undefined subject of supervision responsibility, weak laws and regulations, and lack of supervision system). The findings of this study can help policymakers develop effective regulations for UAV operations to improve public safety.

Feasibility of Child-Resistant and Senior-Friendly Press-Through Packages: Potential of Different Materials

Press-through packaging (PTP) is the most common type of drug packaging in Japan, and a production procedure for PTP has been established at an acceptable cost. However, unknown problems and new needs with regard to safety among users of various age-groups still need to be examined. Considering accident reports involving children and older adults, the safety and quality of PTP and new forms of PTP, such as child-resistant and senior-friendly (CRSF) packaging, should be evaluated. We conducted an ergonomic study on children and older adults to compare types of commonly used PTP and new varieties of PTP. Opening tests were attempted by children and older adults using a common type of PTP (Type A) and child-resistant (CR) PTP (Types B1 and B2) made from soft aluminum foil. The same opening test was conducted on older patients with rheumatoid arthritis (RA). The results showed that CR PTP was difficult for children to open: only 1 out of 18 children could open Type B1. On the other hand, all eight of the older adults could open Type B1, and eight patients with RA could easily open Types B1 and B2. These findings suggest that the quality of CRSF PTP can be improved with the use of new materials.

An integrated EDIB model for probabilistic risk analysis of natural gas pipeline leakage accidents

Natural gas pipeline construction is developing rapidly worldwide to meet the needs of international and domestic energy transportation. Meanwhile, leakage accidents occur to natural gas pipelines frequently due to mechanical failure, personal operation errors, etc., and induce huge economic property loss, environmental damages, and even casualties. However, few models have been developed to describe the evolution process of natural gas pipeline leakage accidents (NGPLA) and assess their corresponding consequences and influencing factors quantitatively. Therefore, this study aims to propose a comprehensive risk analysis model, named EDIB (ET-DEMATEL-ISM-BN) model, which can be employed to analyze the accident evolution process of NGPLA and conduct probabilistic risk assessments of NGPLA with the consideration of multiple influencing factors. In the proposed integrated model, event tree analysis (ET) is employed to analyze the evolution process of NGPLA before the influencing factors of accident evolution can be identified with the help of accident reports. Then, the combination of DEMATEL (Decision-making Trial and Evaluation Laboratory) and ISM (Interpretative Structural Modeling) is used to determine the relationship among accident evolution events of NGPLA and obtain a hierarchical network, which can be employed to support the construction of a Bayesian network (BN) model. The prior conditional probabilities of the BN model were determined based on the data analysis of 773 accident reports or expert judgment with the help of the Dempster-Shafer evidence theory. Finally, the developed BN model was used to conduct accident evolution scenario analysis and influencing factor sensitivity analysis with respect to secondary accidents (fire, vapor cloud explosion, and asphyxia or poisoning). The results show that ignition is the most critical influencing factor leading to secondary accidents. The occurrence time and occurrence location of NGPLA mainly affect the efficiency of emergency response and further influence the accident consequence. Meanwhile, the weight ranking of economic loss, environmental influence, and casualties on social influence is determined with respect to NGPLAs.

Investigating the Work-related Accidents in Iran: Analyzing and Comparing the Factors Associated With the Duration of Absence From Work

Objectives: Work-related accidents as one of the most prominent indicators of safety and health in the industry play an undeniable role in developing and improving the quality of industrial health management. Therefore, their accurate analysis on a large scale is essential. To provide controlling solutions, we aimed to investigate the causes of occupational accidents in Iran for 10 years (2007-2017). Methods: At first, the reports of 10-year work-related accidents were taken from the Social Security Organization, then classified, and entered into SPSS software, version 18.0. Analyses were performed in two stages. Descriptive analyses were initially performed and then multivariate Cox regression was implemented to determine the significance of the effects of the identified factors. Accordingly, six parameters including the cause of the occurrence, gender, type of insurance, time of the accident, marital status, and type of accident were extracted as the effective factors. Next, the effect of each parameter was examined using multivariate Cox regression with a 95% confidence interval. Results: The findings showed that falls and slips (18.3%) and physical strikes (14.6%) were the most common causes of accidents, and carelessness (61.9%) and equipment density (21.1%) had the strongest impact on accidents. Accidents had the highest incidence rate from 9 AM to noon (34.77%). The findings also showed that accident location (P<0.001), cause of the accident (P<0.001), type of insurance (P<0.001), and age (P<0.001) had significant effects on the duration of absence from work. Moreover, the findings demonstrated that the mental condition of workers and workplace design/layout had the strongest impacts on the rate of accidents. Discussion: to control and reduce the risk of occupational accidents, appropriate working situations should be provided.

Application of fuzzy DEMATEL approach in maritime transportation: A risk analysis of anchor loss

Vessels are exposed to various operational risks throughout their service life. Anchor loss, which is one of the operational risks, may lead to grounding, collision and even total loss of vessel afterwards. Accidents due to loss of anchor may lead to serious consequences such as pollution, injury, loss of life and property. Therefore, it is very important to minimize the risk of anchor loss. In this study, a comprehensive literature review including accident and incident reports, technical reports and expert views was conducted to determine the factors that may cause anchor loss of ships. In order to analyze and define the cause-effect relationship between the determined factors, combination of the fuzzy logic and the DEMATEL (Decision Making Trial and Evaluation Laboratory) method was used. The aim of this study is to mitigate the risk of anchor loss by developing an alternative risk analysis tool. Environmental factors at anchorage area, training and competency of crew, the number of shackles paid out, anchoring method and selection of anchor are identified as the most critical factors for anchor loss. According to the results, a risk analysis model based on cause-effect relationship between the factors has been presented for the shareholders in maritime industry.

Analysis and assessment of the human factor as a cause of occurrence of selected railway accidents and incidents

Abstract Ensuring safety is the key factor in the development of rail transport. For this reason, the railway market is heavily regulated, including recording and analysis of railway accidents and incidents. The article presents analysis of 148 final reports of accidents and incidents that occurred in the year 2020 in the Polish railway network and whose direct cause was human factor. The analysis looks at the age and experience of railway employees, as well as analyses human error using the Human factor analysis and classification system (HFACS) model and included train drivers and infrastructure-related stuff. It is pointed out that the available publications are generally focused on signal passed at danger accidents and incidents, which relate to passing through a stop signal, or another stop location, most often caused by an error of the train driver. Therefore, these analyses ignore accidents and incidents related to an improperly prepared and protected route, e.g. failure to close the level crossings boom barriers or incorrectly placed junction point, which under minimally different circumstances could have led to a tragedy. The causes of such accidents and incidents are also related to the human factor and therefore should have been identified and classified. The article identifies weaknesses in the investigations, including the quality of the documentation prepared. Such documentation most often focuses on the factors and causes directly related to the site of the accident or incident, while neglecting the factors at the organisational level.

ASSESSMENT AND PERCEPTION OF OCCUPATIONAL RISKS IN WASTE PICKER ORGANIZATIONS: A PORTRAIT OF WASTE PICKERS SITUATION AFTER FORMAL INTEGRATION

Waste picker organizations (WPO) are a fundamental link in the integrated management of urban solid waste; however, despite being formally recognized, waste pickers still face unhealthy work conditions. Studies on occupational risks related to waste picker activities have been carried out in a qualitative way, but the quantification of occupational risks is an important research gap to fill. Additionally, an unprecedented comparison between waste picker risk perception and occupational safety technician risk assessment is presented. The risk perception of the waste picker was carried out through a cross-sectional interview study that surveyed 35 WPO by non-probabilistic sampling, and the results showed that waste pickers underestimated the occupational risks (i.e. noise, physical effort, improper physical arrangement, and inadequate use of personal protective equipment (PPE)), it was also possible to identify the necessary strategies to improve occupational safety. Occupational safety technician evaluations were carried out through quantitative analysis on site in 64 WPO. The results indicated the predominance of maximum risk intensity (Level 3 – from a scale of 0 to 3) for biological risk, physical effort, excessive pace, improper physical arrangement, and inadequate use of PPE in all operational activities. The main interventions should focus on implementing Work Accident Reporting, rearranging WPO layout, routinely providing information about importance of PPE use, and continuously developing WPO standards with periodic evaluations of occupational risks using a fractional scale.