Human Error In Maintenance Computer Science Essay

Human Error In Maintenance Computer Science Essay Human errors maybe categorized under six groups with respect to engineering: design errors, assembly errors, installation errors, inspection errors, operating errors and maintenance errors. Maintenance error is the result of the wrong preventive or repair actions carried out by the craft or assigned personnel. Base on the frequency of maintenance being performed the probability of human error increases. To reduce some human errors maintainability is important in the design phase of a component or system. Maintainability according to Ebleing is defined as the probability that a failed system or component will be restored or repaired to a specified condition within a period of time when maintenance is performed in accordance with prescribed procedures. Dhillon describes maintainability as the measure taken during the development, design and installation of a manufactured product that reduces the required maintenance, man hours, tools, logistic cost, skill levels and facilities and ensures that the product meets required intended use. The interactions with humans during the design, installation, production and maintenance phases are extremely important. In some phases these interactions may vary depending on the product, but are subject to deterioration due to human error. With respect to engineering products, Meister describes human error as the failure to carry out a specified task or forbidden action that could result in disruption of scheduled operations or damage to property and equipment. This paper looks at the human errors in maintenance practises. 2.0 General Human Factors in Maintenance Frederick Taylor was one of the earlier people to studied human factors attempting to improve design and increase productivity. In maintenance, systems may fail for numerous reasons, one key element being human factors and errors during the design phase. According to Nertney Et al, the following are a list of different human behaviours: People have a tendency to use their hands for examining and testing People get easily confused with unfamiliar things People are too impatient to take the appropriate amount of time for observing precautions People become accustom to certain colours having certain meanings People sometimes overestimate short distances and underestimate horizontal or large distances People may become complacent and less careful after successfully handling hazardous items over a lengthy period People tend to estimate speed or clearances poorly People responds irrationally in emergencies Instructions and labels are read incorrectly or overlooked Attention is drawn to loudness, flashing lights, bright and vivid colours People have little knowledge of their physical limitations People fail to recheck work for errors after performing a procedure They are reluctant to admit errors or mistakes and they do not see objects clearly People get distracted by certain aspects of a product feature People usually expect valve handles and faucets to rotate counter-clockwise for increasing flow of a liquid steam or gas People carry out task (maintenance) while thinking about other things http://www.iru.org/pix/irupolicies/graph-etac-en_1.gif Figure 2.1 Human Senses and Capabilities The five human senses are touch, smell, taste, visual and hearing. In maintenance, visual, smell, touching and hearing are the more common of the human senses being utilized. 2.1.1 Hearing The ability of human to hear is an important factor in maintenance work. When designing for maintainability, excessive noise may lead to problems such as the need for intense concentration, a reduction in workers efficiencies, lost of hearing, unwanted effects on the tasks being performed. 2.1.2 Sight Sight is stimulated by electromagnetic radiation of certain wavelengths. This is measured on the electromagnetic spectrum for visible light. The human eye is sensitive to greenish-yellow light and it sees it differently from different angles. The following are some facts about the human eye: Normally, the eye can perceive all colours when looking straight ahead. As the viewing angle increases, colour perception decreases significantly. When there is poor lighting it may be impossible to determine the colour of a small point of light source. The source usually appears white. When designing components, designers should consider colours such that weak colour people do not get confused. An example is to use red filter with a wavelength greater than 6,500 Ã…, and try to avoid placing too emphasis on colour when critical tasks are to be performed by tired personnel. 2.1.3 Touch Touch complements a human ability to interpret visual and audio. Touch maybe used to relieve the eyes and ears of the load. The sense of touch has been used for centuries in technical work. For example, touch has been used for detecting surface irregularities and roughness. 2.2 Ergonomic Principles The subject addresses human to system interface issues also called human engineering or engineering psychology. In addition to considering the human to system interface, ergonomics is also concerned with: Human dimension The working environment and its effect on humans Effects of systems on humans http://www.omron-ap.com/technical_guide/safety_component/safety_component_12.gif Figure 3.0 Maintenance Errors Human errors in maintenance occur for a number of reasons. In figure 3, a list of different reasons is shown for errors with respect to maintenance. Studies have shown a positive correlation between task performance and amount of time in a career field, ability to handle responsibility, morale and years or experience. Workers who possess these skills are ranked higher and works are generally done faster and with fewer errors. Figure Reason for Human Errors 3.1 Inadequate training and Experience Workers who are inadequately trained and lack the experience on a machine or component will most likely make errors when performing maintenance tasks. They will spend additional time trying to understand what to do rather than effective execution of the task. Errors such as correct shut down procedures, tightening of bolts, correct alignment of components may be done wrong and adversely affecting production, increasing maintenance cost as maintenance work may need to be repeated and can be a safety risk to the maintenance personnel, operator, or the equipment. 3.2 Poorly Written Maintenance Procedures Due to poorly written procedures, there could be cases of components going in wrong place or new components being incorrectly installed. This as with the previous reason can cause harm to craft personnel, operator or the machine. Tasks will take longer because the craft personnel following the procedure will have to spend more time figuring out what to do rather than carrying out the task in a timely and effective manner. 3.3 Fatigued Maintenance Personnel A fatigued maintenance personnel is a receipt for disaster. The personnel will lack concentration and will miss minor details of the work. There will be general tendency to rush the job in attempt to finish fast increasing the probability of errors. This may be influenced by the time of the day, stresses (physical, drug, social, personal) http://elsmar.com/Error_Proofing/img014.jpg Figure 3.4 Complex Maintenance Task Complex and tedious maintenance task may be subject to human errors because of the high levels of concentration required. Humans usually have a low attention span and works requiring excessive hours of concentration may have errors induced. Such jobs should be done in teams so that if one personnel makes an error it can be corrected by other members on the job. 3.5 Improper Work Tools A job requiring a specific tool should not be substituted with an inferior one. An example of this is when a specified torque may be required on a bolt or nut requiring a torque wrench or spanner to achieve a level of precision and accuracy and a regular wrench or spanner is used. This may cause misalignment, et cetera. 3.6 Poor Work Layout Poor work layout is inefficient and can cause errors to be made by the maintenance personnel. 3.7 Poor Work Environment A poor work environment will cause discomfort to the maintenance personnel such as humidity, lighting and heat. Concentration levels will be low and there will be a rush to complete the task. 3.8 Poor Equipment Design In some equipment, components may be difficult to reach to perform the necessary works. Errors may be caused in the process as the tasks may not be carried out completely. 3.9 Outdated Maintenance Manuals It will be extremely difficult for maintenance personnel to follow out-dated maintenance manuals. In such an instance, the procedure will most times be incorrect inducing a number of errors in the job. 3.10 Factors Contributing to Human Error Reason and Hobbs have studied and researched the psychological and physiological factors that contribute to inevitable human error. These will include: Differences between the capabilities of our long term memory and our conscious workspace. Attention span is limited if it is focused on one thing. Also we can only attend to a very small proportion of the total available sensory data we receive Unrelated matters tend to capture our attention Concentration is hard to maintain for lengthy periods of time The ability to concentrate depends strongly on intrinsic capability of the current object attention Habitual actions are done will less attention than normal The right balance of attention is required to correctly perform a task The vigilance decrement- inspectors miss fault because it may become redundant after doing it for a long period The level of arousal- too much or too little impairs work output Biases in thinking and decision making Confirmation bias where we seek information that confirms initial diagnosis of a problem Emotional decision making is when a situation is frustrating then an aggressive approach is induced http://www.lifetime-reliability.com/images/wrongs-humans-do-to-machines.gif Figure Human Maintenance Errors Base on the aforementioned factors contributing to maintenance errors, the following are the most common types of errors: Recognition Failures These include misidentification of objects, signal and messages and non detection of problem states. Memory failures Input failures- Insufficient attention is to the to-be-remembered item. Storage failures- remembered material suffers interference Output failures- things that we know cannot be recalled at the required time Omissions following interruptions Premature Exists- End a job before all tasks are complete Skilled-based Slips Usually associated with routines and they can include: Branching errors- an error made based on a custom and not knowing when to deviate. Over shoot Errors- having an intention but forgetting to do it Rule base Mistakes Misapplying a Good Rule- using a right rule in a wrong situation Applying a bad rule- unwanted consequences in using a rule even though the job will be done. Knowledge Based Errors It is common when someone is doing a task for the first time but not always the case. Violations Purposeful acts which violates procedures. These may be: Routine violations- done to avoid unnecessary effort, gets the task done quickly, to demonstrate skill or avoid what is seen as an unnecessary laborious procedure. Thrill seeking violations- often committed in order to avoid boredom or win peer praise Situational violations- these exist because it is not possible to get the job done if procedures are strictly adhered to. Figure Summary of main error types 4.0 Mitigation/Reduction of Hum Errors in Maintenance It is impossible to total alleviate all human errors when it comes to maintenance, but Engineers and designers have worked together and is continuing to find ways to reduce some human errors. However a lot of the responsibility is on the maintenance personnel to ensure that maintenance tasks are carried out effectively. This chapter looks at ways to reduce human errors in maintenance. 4.1 Avoid Unnecessary Preventive Maintenance Over maintaining equipment, not only wastes time and money but it also increases the risks of environmental incidents but also causes expensive and unnecessary failures. Reliability Centred Techniques such as Failure Mode and Cause Analysis, Fault Tree Analysis and others can be useful eliminating unnecessary maintenance works and optimizes and streamlines the equipment preventive maintenance. 4.2 Standardization Ankenbrandt et al noted that standardization is the attainment of practical uniformity in product design. Non-standard equipment or component lessens the reliability and increases maintenance. Standardization restricts the number of components equipment will require. The advantages of standardization according to Dhillon are: Reduction of using the incorrect parts Reduction in wiring and installation errors due to the fact of variations in characteristics of similar items Elimination for the need for special parts Reduction of incidents because of the use of wrong or unclear procedures Reduction in design time, manufacturing cost, and maintenance cost Reduction of procurement, stocking and training Figure 7 shows the goals of standardization. Figure Goals for reducing errors 4.3 Modularization Modularization deals with a production into physically and functionally distinct units to allow removal and replacement. The following are guidelines for the usefulness associated with modular units: Aim to make modules and parts as uniform as possible with respect to size and shape Divide the equipment or item under consideration into many modular units Aim to make each module of being inspected independently Design the equipment so that a single person can replace a failed part with ease and also parts should be small for mobility Module should be designed for maximum ease of operational testing when it is removed from the actual equipment or system. Consider design, modularization and material problems simultaneously For ease of disconnection, design components with control levers and linkages to allow easier replacement. 4.4 Simplification and Accessibility Design engineers should consider simplification in the design phase and it should be constant. Consideration should also be taken with the important functions of a system or a product into the design itself. Accessibility is the ease with which an item can be reached for repair, replacement or servicing. Poor accessibility will result in sub-standard maintenance. Accessibility is affected by: The visual needs of personnel performing maintenance task, The location of an item and its environment. Distance to be reached to access the component or part The types of maintenance to be performed through an access point or opening The danger associated with use of an access opening The clothing worn my maintenance personnel The task required time for execution The types of tools and accessories required to perform task Mounting of items behind the access point Work clearance to carry out task 4.5 Interchangeability and Identification Interchangeability refers to as an intentional aspect of design, that any component can be replaced with a similar item. This is achieved through standardization. There are three basic principles for interchangeability according to the Department of Defence (DOD): In items, components and products requiring frequent servicing, replacement of parts, each part must be interchangeable with another similar part Liberal tolerance must exist Strict interchangeability could become uneconomical in items that are expected to operate without replacement. The following considerations are taken to achieve maximum interchangeability of parts: Existence of functional interchangeability when physical interchangeability is a design characteristic Sufficient information in task instruction and number plate identification should be available for allowing users to decide confidently whether two similar parts are interchangeable No change in method of connecting and mounting when there are part or unit modifications Avoid or stay away from differences in size, mounting and shape Availability of adapters for making physical interchangeability possible when total interchangeability is not practical There should be and identification system for total interchangeability of identical parts 4.5.1 Identification Identification is concerned with labelling or making of parts, controls and test points to facilitate tasks such as repair and replacement. Maintenance task becomes more difficult when parts and components are no properly identified. It usually takes longer and increases the risk of human induced errors. Identification could be for parts or component or equipment identification. 4.6 Task Analysis The task analysis is a formal methodology derived from systems analysis which describes and analyse performance demands made upon humans within a system. The aim is to achieve integration of humans and machine system components. 5.0 Assessing the Risk in Maintenance Risk assessments of all human activities have great importance for the prevention of major accidents. The risk with respect to human errors when dealing with maintenance needs to be assessed and rank based on the overall impact the risk can have on the operation of the organization. The Risk Management Framework used is shown in Figure 8.0 This program was utilized to conduct a risk assessment and evaluation before and after the risk control measures have been put in place. The main aspects of the Risk Analysis procedure are: i. Identify Potential Hazards and existing risk control or safeguards. ii. Estimate the consequences of the potential Hazard. iii. Estimate the frequency of each of the impact/consequence of the potential hazard. iv. Determine the Risk without considering any new risk control or safeguards v. Determine the Risk with proposed new safeguard vi. Evaluate Economic impact of proposed new safeguard vii. Optimize and propose additional modification. Figure Risk Management Framework As indicate in the previous chapter (3), each human error identified will have to be ranked via a metric scoring system to determine the hierarchy of the error. The impact of these errors will have to be taken into consideration with respect to the employees (operators and maintenance staff), the equipment and the environment. Figure Diagram Showing Contributors of Risk 6.0 Human Errors in Maintenance in Industry Human error in maintenance exists in some form or the other in all industries. However, in some industries the consequences of human error from maintenance task can be catastrophic. Industries such as nuclear plants and civil aviation cannot afford errors. 6.1 Human Maintenance Errors in Power Generation Human factors play an important role in power plant maintenance because improving the maintainability design of power plant facilities, systems and equipment assists to improve direct and indirect plant productivity, availability and safety. According to Seminara power lost can cost between US $500,000 to 750,000 dollars a day. The following are human related deficiencies in maintenance of a power plant: Limited access or inadequate clearance to perform maintenance Equipment poorly designed to facilitate maintenance activity effectively Equipment Inherently unreliable Personnel safety hazard Impaired mobility for equipment and personnel Miscellaneous- lack of standardization, high temperature environment and poor air conditioning While maintenance activities of nuclear power plants are essential for sustaining the safety of a power plant and maintaining the reliability of plant systems and components, they also have potential of human errors leading to unplanned reactor trips or power derate (Dhillon 2006). J. Reason classifies in chapter 3 the different characteristics of human errors that will be applicable to Power Plants. According to Kim et al, incidents reports in Korea stated that most of the human related unplanned reactor trip events during the normal power operation are associated with test and maintenance activities (63%). Plant maintenance included preventive maintenance, planned maintenance and corrective maintenance. 6.2 Human Error in Aviation Due to the increase in air travels, airlines spend billions of dollars in maintenance annually. A study done in the United Kingdom has shown that between 1990 to 2000 maintenance errors per million flights has doubled (ICAO) According to the ICAO there are over 300 factors and influences that can impact the performance of maintenance personnel. Some of the human errors that are impacted by maintenance are listed in chapter 3 and range from time pressure, inadequate training to outdated maintenance manuals. There are many categories of human errors with respect to aviation maintenance and inspection. Some of the major ones encountered according to Latorella et al are: Incorrect assembly sequence Procedural defects Wrong part Incorrect configuration Missing part Defective part Functional defects and tactical defects Figure List of Errors in Aircraft Maintenance Figure 10 shows the guideline covered in aviation. Two important guidelines concerning design are: Seek relevant information on human error occurrence during the maintenance phase as inputs in the design phase Ensure that manufacturers give attention to maintenance related human factors during design phase According to James Reason based on a Boeing study which indicated the top seven causes for in-flight engine shut downs on the Boeing aircraft were as followed: Incomplete Installation (33%) Damaged on installation (14.5%) Improper Installation (11%) Equipment not installed or missing (11%) Foreign Object Damage (6.5%) Improper fault isolation, inspection and test (6%) Equipment not activated or deactivated (4%) From the statistics presented only one cause was unrelated to maintenance activities. Maintenance activities were responsible for approximately 80% of in-flight engine shut downs. 7.0 Prediction Models for Human Errors Human error can be analysed by several methods developed over the years. Some of the following methods are relied on heavily in the aviation industry. 7.1 Cause and Effect Diagrams This is a useful tool to determine root causes of specified problem in relation to maintenance errors. Related and relevant ideas are generated. The extreme right of the diagram represent effect and to the right are all the possible causes. The cause and effect can be generated in the following steps: Develop problem statement Brainstorm to identify possible causes Establish major causes categories by stratifying into natural and process steps Connect diagram to all the causes by following the appropriate process steps and filled in the effects Refine cause categories by asking What cause this and Why does this condition exists? 7.2 Error Cause Removal Program (ECRP) This was developed to reduce the occurrence of human error to a tolerable level in production operation. It focuses on preventive measures rather than remedial ones. It is composed of a team of workers with each team having its own coordinator with the necessary skills. Each team member presents a report and the coordinator discusses these reports in a periodic meeting and recommendations are made to management. 7.3 Fault Tree Analysis The fault tree analysis (FTA) is used to perform reliability and safety analyses of engineering system and can be use to perform analysis of human error in maintenance. 7.4 Markov Analysis This method is used to perform reliability analysis of engineering systems and can be used to predict the probability of occurrence of human error in maintenance. This is done through mathematical modelling. 7.4.1 Model 1 This mathematical model indicates a system that can fail due to maintenance error or other failures. See figure 11. Numerals in box, circle and diamonds represents system states. The following assumptions are made for this model. The system can fail due to maintenance error or failure other than maintenance error The failed system is repaired and maintenance is performed periodically Fail system repair rates are constant The repaired system is as good as new Figure Markov Model 1 Diagram 7.4.2 Model II This mathematical model represent a system that can only fail due to non maintenance related failures but its performance is degraded by the occurrence of maintenance error. See figure 12. The numeral in circle, box and diamonds represents system states. The following assumptions are made for this model: The total or partially failed system is repaired and preventive maintenance is performed regularly The occurrence of maintenance error can only lead to system degradation but not failure The system can fail from its degradation mode due to failures other than maintenance errors The system is repaired at constant rates from its failed and degradation states Maintenance error and non maintenance error failure rates are constant The repaired system is good as new Figure Markov Model 2 8.0 Conclusion This report gives a comprehensive analysis of human errors in maintenance. Human errors are inevitable and there are numerous factors that can influence these errors. The first step is to identify the different human errors that may be encountered in an industry, conduct an assessment on the errors and take necessary steps to reduce these errors. The traditional approach of dealing with human error, that is counselling and or re writing procedures may not be effective in dealing with the errors identified in this report. A more holistic approach for managing maintenance error and assuring maintenance quality techniques is the application of reliability centred maintenance and also designing for maintainability.