Abstract of Special Research Report (SRR-No.28)

National Institute of Occupational Safety and Health, Japan

A Study on the Prevention of Accidents Caused by Human Errors in the Construction Work (Midterm Report)



: The death toll of the labor accident decreases gradually for the long term. However, the decrease rate shows the tendency to become duller recently. In the background, some factors might exist; 1) As the working environment has been improved than before, it would be difficult for the workers to recognize the risk in the working environment. 2) As the workers should do more advanced and complicated work, the hazard of the disaster become more complex. 3) As the number of experienced workers has decreased gradually, it would be difficult to follow the knowledge and the technique concerning safety management. 4) As the workers have less experiences of the disaster, their sensitivity to risk would become dead. From the end of past century, the social and economical environment which surrounds workers have changed greatly, and the characteristics of the labour accidents might change more and more, too. It should be necessary to make a new paradigm in the approach of the labor accident prevention in the near future.
    About the labour accidents, the fatalities in the construction industry account for about 40% constantly. The disasters caused by human errors are serious, because the automation in the construction is difficult and a lot of workers support the construction work. The larger the number of workers who work on the site is, the higher the possibility the human errors cause the accident is.
    This project consists of a few studies;
    First, the safety attitude of workers and environmental factors of construction work sites, which are regarded as the cause of human errors and risk taking, are examined by means of field research such as hearing, questionnaire, and so on.
    Second, the most of human errors are related to the cognitive process. Especially, visual information is very important for human. So, the occurrence mechanism of human errors is investigated in some experiments which attach importance to psychological aspects, such as the abilities to detect the targets (sometimes they are dangerous ones), risk-sensational sphere of workers, and so on.
    Third, the occurrence mechanism of human errors, which is connected with the communication about safety information, is investigated. One of the influential factors for human errors is the interaction among human beings such as the contents of information, the way to communicate and the situation of conveyance.
    In addition, the aging is an important problem even in the construction industry. Therefore, the interest of the physiological-psychological workload and human errors are examined from a view of the aging.
    Based on these studies, this project proposes the desirable methods of safety education and safety training. Their purpose is the prevention of a disaster caused by human errors or risk-taking, and these supply a new technology for the safety management in the construction industry.

A Survey on the Unsafe Behavior and its Prevention Measures at Construction Sites

Takuro SHOJI and Yoshiyuki EGAWA

: Due to the efforts of researchers, companies, and government resulting in advances in safety technology, improvement of the working environment, development of machines, and the various countermeasures of the industry, the number of labor accidents in the construction industry in Japan has dropped to lower than in previous years. However, the speed of reduction has decreased in the last several years. Recently, accidents caused by human error or unsafe behavior have come to make up a large portion of all the labor accidents.
    In order to decrease accidents further, much attention has been paid to the human factor and human behavior in addition to the adjustments.
    The purpose of this study is to clarify the occurrence, cause and prevention of unsafe behavior at construction site by interview, literature review and questionnaire survey.
    By the interviews to employees of general contractors and foremen of affiliated companies and literature reviews, it was revealed that more accidents at work sites were caused by workers' unsafe behaviors rather than malfunction of machines or imperfect equipments. And the conceptual structure model consist of six causes of unsafe behaviors, i.e. workers' factor, work situation factor, work environment factor, management factor, organizational factor and the other factors was constructed.
    Concrete items that cause unsafe behaviors are revealed by the questionnaire as, "in a hurry", "bad safety equipment at the site" and "untidy work place".
    On the other hand, what seem to be effective to prevent unsafe behaviors are "keep work place tidy", "small group activity for safety at work site like KY activity", "well equipped work sites" and "lively activity of foreman association".
    A group of workers called "foreman association" was formed on almost all the large work sites and its activity had a great effect not only on promoting friendships among workers but also on promoting atmosphere in which workers would be able to warn unsafe behaviors to each other.
    But there was difference of opinions among office staffs, employees of general contractor and foremen of affiliated companies, about the effect of the prevention methods of unsafe behaviors.
    Further survey will help to modify conceptual model taking more variables, such as work site climate and workers' attitudes, and also help to clarify what is effective to make workers behave safely and to make construction work safe.

Changes of Working Behaviors by Spreading Safety Information in Construction Sites

Tsuyoshi SAITO and Hiroyasu IKEDA

: In order to investigate into the change for the safety behavior of workers by spreading safety information in construction sites, the survey and experiments were carried out. These outlines about the survey and experiments were as follows.
    To prevent unsafe behavior due to a lack of the safety information in construction sites, the influences for the contents of the information from the interviews with the director of the site, the foremen, and the workers were surveyed. Further, the survey for the circumstances of informational meetings, the contents of the information and the channels of the information were conducted. In construction sites, the circumstances of informational meetings were morning assemblies, working adjustment meetings, educations of new attendant workers, KY activities, conferences for safety & health. The channels of the information could be roughly classified into two types. One type was "Lengthwise Channel Type" which was a formal one-way communication from the director of the site to the workers in the strata of construction work. The other type was "Evenness Channel Type" which was an informal bi-directional communication by workers on flat in the construction strata. It was possible for "Lengthwise Channel Type" to inform many workers. However, the spreading much information by using a one-way communication came into the many workers who didn't pay attention. And the speeches for many workers by using the one-way communications made a vague contents like "Pay attention!" or "Keep to the safety regulations!".
    Experiments were conducted to invest the relation between the safety behavior and the safety information. The experiments were conducted under two conditions. One condition was the way of the enforcement of the safety regulation, and the other was the way of the presentation of danger scene by the pictures. The experimental condition of the enforcement of the safety regulation had many lack of care for subjects than presentation for danger scene. However, the safety behavior of the precautions for the surrounding circumstances was increasing under the condition of the presentation for danger scene.

Ergonomics Study on High-Rise Working Conditions in Construction Sites

Yoshiyuki EGAWA, Shin-nosuke USUI, Takuro SHOJI and Takahiro NAKAMURA

: The number of deaths in industrial construction accidents was 644 in 2001. The largest cause was falls, killing 262 workers and constituting 40.7% of all deaths. Fall accidents have constituted around 40% over the last 10 years, and the number is not declining.
    Based on these conditions, ergonomics study on the safety of high-rise working conditions to prevent the falls of workers was conducted. This study consists of two parts. The first part is fatal accidents analysis in the past years and the second is the result of experiments on work load at high-rise working conditions.
<Fatal fall accidents analysis>
    The investigation was performed on fatal fall accident reports. At this situation, 200 case reports for building construction and wooden construction of those workers aged over 50 and those aged under 40 were selected as the objects of this study. Compared to building construction, the workers with longer experience fell to their death more in wooden constructions. Especially, for the workers under 40, the workers had an average of 4 years experience in building construction, and the other hand, for workers of 50 or over, the workers had an average of 30 years experience in wooden constructions. In building construction, the falls occurred during movement on the scaffoldings, on the other hand, the accidents occurred often during working in the wooden constructions.
    The experiment on work load at high-rise working conditions was conducted with 16 subjects. These subjects those were 7 elderly workers and 9 younger workers were walking and carrying the boards on scaffolding. Walking speed, step distance and foot stroke time were recorded by video camera. Heart rates on physiological working load and mental capacity on psychological working load were measured. At the result of this experiment, it was cleared that the high-rise condition of working with the difficult behaviors for elderly workers had increased physiological and psychological working load.

A Study on the Task Performance on the Different Thermal Conditions

Takahiro NAKAMURA, Kiyoshi FUKAYA and Sonoko MANNEN

: As construction works are often operated in outdoors, workers are affected by work environments. Especially in summer, workers suffer from heat stress in addition to physical work load. While the measures of effective prevention against heat stroke have been considered in the industrial hygiene field, prevention of human errors during work in hot environment is not taken adequately. There are no criteria for stopping or continuing work from the viewpoint of safety. But it is known that the task performance deteriorates in uncomfortable environment, and there seems to be the high possibility of occurrence of human errors or accidents during work in hot environment.
    The final goal of this study is to examine the effects of heat stress during construction work on work performance and to clarify the possibility of occurrence of human errors. In this research, as the first step, the effects of temperature of environment on task performance and laps of attention were examined.
    Eight healthy male students took part in the experiment as test subjects. They carried out physical task, consisting of taking apart a rack of shelves, carrying the parts to appointed places, and putting back together for 90 minutes on three thermal conditions, 23 °C, 29 °C and 35 °C. During the physical task, each subject was forced to listen to the sound selected a number out of the group one to nine and when the fixed number was called they were ordered to stop tasks and respond by PC mouse. ECG and rectal temperature were recorded throughout the experiment. Subjective symptoms of physical fatigue were also measured during tasks.
    On 35 degrees condition, averaged frequency of lapse of attention, not responding fixed number, was the highest. On each condition, the number of errors tended to increase as time of task passed. Significant effects of time (0-30 minutes, 31-60 minutes, 61-90 minutes) and temperature (23 °C, 29 °C, 35 °C) were found by the ANOVA. Similar tendencies were also shown for HR, rectal temperature and subjective fatigue.
    The time-series variation patterns of missing the number, task performance for each subject and physiological and psychological states are not identical. So task performance, physiological state and subjective fatigue are not proper prediction measures for errors during simulated work, and workers who are doing their jobs neatly are also in danger of making errors to some extent.
    These results suggest that there is much possibility of occurrence of errors during physical task in hot environment, especially in the case that the work period is long, and that it is necessary to consider some preventive measures to keep safety at work in a hot environment.

Eye Movement and Fixation in the Excavator Operation

Takahiro NAKAMURA, Kiyoshi FUKAYA and Sonoko MANNEN

: The simulator is a remarkable tool for safety education and training, and it is considered to be useful as an experimental approach as well. Nakamura et al. (2001) measured eye movement of excavator operators through the experiment using the simulator. They discussed the fundamental factors to acquire visual information, which affect the useful field of view, such as experiences, mode of the operating levers, and so on.
    However, there is some difficulty to consider the observed behaviour in the experiment using the simulation to that in the real situation. Because the simulation is consists of artificial environment, even if it contains the advanced reality, so the subjects distinguish easily the simulation and the actuality.
    From these points, it is necessary to understand the differences between the result of experiment using the simulator and that using real machinery. Through such examination, it will be possible to confirm the validity of the simulation.
    In this study, eye movement measurement was carried out using real excavator (actual-condition) (N=2). As a contrast, the data of the experiment using the simulator (simulation-condition) was used (N=4). Some fundamental factors to acquire visual information were compared and discussed, such as the extent of eye movement, saccadic distance, fixation duration, frequency of fixation to each object, and so on. As the result, several common aspects were found between both experimental conditions.
    Even if the obtained results were limited to visual information factors, the validity of the simulation would be confirmed, and the utility value of the simulator as an experimental equipment would increase.

Influences of Time-Pressure in the Excavator Operation

Takahiro NAKAMURA, Kiyoshi FUKAYA and Sonoko MANNEN

: In the construction work, there are time related factors such as the insufficient time allowances, the impending of the complete date, and the several stresses in the emergency (recoveries of mistake, weather influences, etc.) and so on. These factors are regarded as the background factors of the occupational injuries, because they might cause the omission, shortening and turbulence of the procedure as the result of making haste, impatience and panic.
    The construction machineries are very useful for labor saving and increasing efficiency, however they often cause the serious injuries because of their huge energy. From these points, the effect of time-pressure about the operation of the excavator was discussed in this study.
    In the experiment, excavator simulator was used. The subjects were six and they had no experiences of operation. After training session, they took part in the experiment which consisted of two conditions, one was the control (non time-pressure), and the other was the time-pressured. The rapidity and accuracy of operation were measured. While operations, the target detection task was applied to the subjects to determine their peripheral vision performance.
    Under the time-pressured condition, the rapidity of the operation was higher than control condition. The differences in terms of the accuracy of operation and peripheral vision performance were not clarified between both conditions.
    From these results, the time pressure to the subjects in the experiment seemed to be not sufficient. The subjects might feel tired to the tasks in the control condition, so their performance of operation and vision was reduced. In the time-pressured condition, on the other hand, the time pressure might be moderate and not enough to make the subjects feel panic, haste and impatience, so the performance of the subjects was promoted.
    The experimental method, which manipulates the time pressure, will have to be dealt with more carefully. Further, it is necessary to understand how the workers are exposed to the time-pressure, how and what stress related time they feel in the actual work sites.

Development of Excavator Accident Simulator

Kiyoshi FUKAYA, Takahiro NAKAMURA and Sonoko MANNEN

: An excavator accident simulator is developed in order to study on the effect of quasi-experiences of accidents. This development is made from the point of view that quasi-experiences of accidents are believed to be effective for safety training for acquiring sense of danger and motivation for safety.
    In this simulator, excavator accidents are selected as quasi-experiences of accidents. They seem to be representative of accidents in a sense. One reason is the number of excavators' accidents is large. Another reason is there are two type of accidents; own-injury accidents and accidents hurting others.
    Hardware of the simulation system is composed of the image system and the motion system. The image system is composed of the image generator (graphic workstation), eight video projectors and eight 120 inch sized screens which are arranged in heptagon and surround a subject. The graphic workstation performs other functions at the same time, such as the system control function and the input-output control function for control levers. The motion system has an operator seat on its motion base and it gives a sense of motion to the operator.
    Software of the simulation system is composed of an application system and the VR(Virtual Reality) operating system. An application system is composed of 3D models and application program. The VR operating system controls drawing of computer graphics according to the application program and 3D models.
    The excavator accidents simulator simulates 4 types of works, that is, excavation, running on sloping ground, hanging a load and loading of the excavator, and it simulates 2 types of accidents, that is, upset and contact.
    In the simulation of upset, stability and posture calculation is carried out. In the running on sloping ground posture of the excavator changes according to the sloping ground. In order to express inclination of the posture, not only screen image but also motion of operator seat is used.
    An experiment for evaluation is carried out, and modifications are suggested. Modifications of the system are being carried out.

Other Publications

Kiyose District Map

Umezono 1-4-6,Kiyose,
Tokyo 204-0024 Japan
TEL: +81-42-491-4512
FAX: +81-42-491-7846

Noborito District Map

Nagao 6-21-1, Tama-Ku,
Kawasaki 214-8585 Japan
TEL: +81-44-865-6111
FAX: +81-44-865-6124