CRE (Hong Kong Civil Service Exams) key terms

The probability that an item performs its required function, under stated conditions, for a stated period of time, expressed as a value between zero and one.

The probability that a failed item can be restored to a specified functional condition within a stated time when maintenance is performed under prescribed procedures and resources.

The fraction of time an item is in a state to perform its required function, combining how often it fails with how quickly it is restored.

The conditional probability that an item fails in the next unit of time given that it has survived to the start of that interval, usually denoted lambda and expressed per hour.

The instantaneous rate of failure at time t for an item still functioning at that time; for a constant-failure-rate model it equals the failure rate.

The expression R(t) giving the probability an item survives beyond time t without failure; it starts at one and decreases monotonically as time increases.

A graphical model of failure rate over product life showing three phases: decreasing infant mortality, a roughly constant useful life, and increasing wear-out.

Early-life failures caused by manufacturing defects, material flaws, or assembly errors; the rate decreases with time and is often removed by burn-in screening.

Failure whose likelihood increases with accumulated age or usage due to degradation mechanisms such as fatigue, corrosion, friction, or insulation breakdown.

A failure whose occurrence in time cannot be predicted, arising from chance overstress or inherent variability, and typically modeled with a constant failure rate.

The probability model for time to failure when the failure rate is constant; it has the memoryless property and a mean equal to one over the failure rate.

A flexible life distribution whose shape parameter indicates decreasing, constant, or increasing failure rate and whose scale parameter sets the characteristic life.

For a repairable item, the average operating time between successive failures; the reciprocal of the average failure rate during that period.

For a non-repairable item, the expected time from start of operation until the first failure, equal to the integral of the reliability function over time.

The average elapsed time needed to restore a failed item to service, including diagnosis, access, repair, and verification activities.

The average total time an item is unavailable, comprising repair time plus administrative and logistic delays such as waiting for parts or personnel.

The average number of repairs completed per unit of time for a repairable item, usually modeled as the reciprocal of mean time to repair.

Availability calculated from design parameters only, using mean time between failures and mean corrective maintenance time, excluding supply and administrative delays.

Availability considering corrective and preventive maintenance time but excluding delays outside the maintenance organization, such as spare-parts waiting.

Availability measured in the field, including all downtime sources: corrective and preventive maintenance, logistics delay, and administrative delay.

A logic model showing how subsystem successes combine to achieve system success, using series, parallel, and k-out-of-n arrangements rather than physical connections.

A reliability arrangement in which all elements must function for the system to function, so system reliability is the product of element reliabilities.

A reliability arrangement in which the system succeeds if at least one element functions, providing redundancy and raising system reliability above any single element.

Providing more than one means of performing a function so that a failure in one means does not cause loss of the function.

A configuration in which redundant elements operate simultaneously and share the load, so all are energized and subject to failure while the function is required.

A configuration in which backup elements are not operating until needed, so they do not accumulate operating stress but may fail to start when switched in.

Failure of multiple redundant elements from a single shared cause such as a common environment, design flaw, or maintenance error, defeating the intended redundancy.

A deductive top-down technique that traces an undesired event through Boolean logic gates to combinations of basic component and human failures.

A smallest combination of basic events whose simultaneous occurrence causes the top event; removing any member prevents that combination from causing it.

A bottom-up team technique that lists each item's potential failure modes, their causes and effects, and ranks them by severity, occurrence, and detection.

A ranking score computed by multiplying severity, occurrence, and detection ratings in an FMEA to prioritize which failure modes need corrective action first.

Ranking failure modes by the product of failure effect severity and failure rate or probability, used to identify which items most threaten mission success.

A technique that searches designs for unintended current paths or logic sequences that cause functions to operate or inhibit when not commanded.

A failure that exists but remains undetected until a demand or test reveals it, often hidden in standby equipment or protective functions.

Hardware and software built into an item that automatically detects and reports faults, enabling faster diagnosis and improved maintainability.

The process of narrowing a detected failure to a specific replaceable item or circuit so that repair can be performed efficiently.

A screening process that operates items under stress for a period before delivery so early-life defects fail in the factory rather than in service.

Applying thermal cycling, vibration, or other environmental stresses to production units to precipitate latent manufacturing defects before field use.

Operating components below their rated electrical, thermal, or mechanical stress limits to reduce failure rates and extend useful life.

A disciplined program of requirements, prediction, part selection, derating, testing, and growth activities that builds reliability into a product during design.

An iterative test-find-fix process in which failures are analyzed and corrected, and the demonstrated mean time between failures rises over the test program.

Testing at stress levels above normal use to shorten time to failure, with results extrapolated to use conditions through a physically based life-stress model.

Scheduled servicing or replacement performed before failure to reduce degradation and avoid the higher cost and disruption of corrective maintenance.

Maintenance triggered by measured evidence of deteriorating condition, such as vibration, temperature, or oil analysis, rather than by calendar intervals.

Using trended condition data and models to forecast when a component will fail, so maintenance can be scheduled just before that point.

Estimating remaining useful life of an item from current condition, usage, and degradation models to support maintenance and operational decisions.

Freedom from unacceptable risk of harm to people, property, or the environment, judged against defined risk acceptance criteria.

The combination of the probability of a harmful event and the severity of its consequences, used to rank hazards and set priorities.

A systematic examination of a system to identify conditions or events that could lead to harm, and to assess and control their associated risk.

A discrete target level of risk reduction assigned to a safety instrumented function, defined by required failure measures such as average probability of failure on demand.

An engineered set of sensors, logic solvers, and final elements that brings a process to a safe state when specified hazardous conditions are detected.

A periodic test that reveals hidden failures in a safety function so that its assumed reliability and integrity level remain valid between test intervals.

A design principle in which a failure causes the system to move to a state that is safer than continued operation, such as de-energize to trip.

A method for estimating the probability and causes of human errors in operating, maintaining, or responding to a system, and for reducing them.

The likelihood that a specified human action fails to be completed correctly within a required time and context, used in system risk models.