The application of reliability to product and process engineering has shown great results, since it is a means to anticipate failures and also know the probabilities of these to occur, it is a means to develop products and processes that are capable to be able to be manufactured; since many of the problems that the production area presents can be prevented through reliability techniques, and thus be able to obtain products according to the customer's expectations.

The branch of reliability as an application was born at the end of the 40s, after what was the industrial revolution, the purpose of this methodology was used to estimate the number of spare parts necessary for the maintenance of electronic and mechanical equipment that were operating intermittently for very long periods of time during the war. Later in the 50s with the development of equipment for space, it was necessary to address more about this issue. In the 70s, in which the oil crisis was found, it marked the beginning of a global crisis and the beginning of leadership by Japan and thus the reliability of products and services.Today this issue is marked by globalization and constant pressure on companies to minimize their production cycles, reduce costs and improve quality and reliability, which is why, together with administration, engineering and statistics, reliability has a great challenge for the years to come.

The concept of reliability can be defined as the probability that a product unit performs satisfactorily, fulfilling its objective during a time that has been designed and under the conditions that were specified. (Acuña, 2003)

A definition from the point of view of quality is that it is maintained over time, so that a reliable product or service remains within its specification limits during its technological life. Not always that quality is not always enough to guarantee good reliability and this can only be evaluated after several uses.

Within reliability are two areas: repairable systems and replaceable components or units. For example, for the first one, the required data describes the trends and failure patterns of a complete system, which are supported by special statistical tools, in this a set of repairable units is monitored where the event of interest may be the failure of the units, repair cost, or both. For the second, the data of components or replaceable units describe times of failure or degradation of units that are repaired, the sources for these data are laboratory tests of materials or components and data of components and replaceable subsystems obtained from tests by monitoring systems.

Reliability data is typically censored due to time or resource constraints.

A graphical and non-parametric analysis of the data must be carried out. The values obtained are positive, for its modeling a non-normal distribution is assumed, but in the same way it is common to use distributions such as the lognormal, the Weibull, those with extreme values, the range, the Poisson. For its analysis there is a combination of graphical methods such as non-parametric and maximum truthfulness estimates, in order to fit non-parametric models. Least squares methods are not inefficient for estimation as well as inference and forecasting. Some of the metrics used with failure rates, quantiles, failure probabilities, reliabilities, and recurrence rates. The mean and variance are of little interest.

Operational reliability

Operational reliability is the ability of the company, through its various processes, technologies and human capital, to fulfill its purpose within the design limits and operational conditions. This is operational reliability. It comprises a series of continuous improvement processes that systematically annex diagnostic tools, methodologies, management, planning, implementation, and control, with respect to industrial production, supply and maintenance.

To achieve the reliability of the operations, an integration of the assets must be achieved, speaking from the design part to the operation and human capital, in the same way the components that structure it must be in an integrated way to generate the reliability of the processes.

The operational reliability has five axes, which together manage to obtain reliable facilities in the long term, these are; the human reliability, maintainability and reliability of the assets that are directly related to the design of the equipment and the logistics part, the reliability of the process and the procedures used in the operation of the facilities, this is to respect the established conditions and reliability of supplies.

In the design of industrial projects the efforts of the engineering part are aimed at optimizing the different operational processes that are in the value chain. Reliability engineering tools allow to calculate in a robust and intrinsic way to each project the probability of failure or non-operation of the production system, in order to establish an input function that forces to optimize the processes evaluating their operational reliability, in order to This way, estimate the real profitability of the project and thus reduce uncertainty to its maximum.

Reliability engineering makes it possible to determine, through quantitative and qualitative support, the solutions at the project level through maintenance plans and continuous improvements that optimize the management and maintenance of assets in order to improve company results. (Arata, 2009)

Among the main aspects of the reliability theory are:

Collection of data with statistical bases that allow product life tests to be carried out and thus be able to determine the reliability of a product or process Selection of the best reliability analysis method that adapts to the analysis and testing requirements Understanding of the concept of Reliability based on the properties of the materials. Analysis of the principles to be able to carry out the program of reliability and product safety.

Reliability engineering tools

FTA

It is a deductive method that has excellent characteristics when it comes to locating, correcting and being able to anticipate faults. First, you must define an event that you do not want, the failure you want to avoid and starting from the identification of the factors that can carry out this phenomenon. It is represented graphically through the well-known fault tree, and the combinations through logical operators and algebraic operations are defined in Boolean. The FTA calculation algorithm is as follows. (Tamasa, Santiago, & Mauricio, 2009)

Define the system to study

Define the unwanted event for analysis (total stop)

Define the tree and its scope

Solve the fault tree

Quantitative analysis

One of the tools used in reliability engineering is reliability and failure mode analysis (FMEA), which is a systematic methodology used to analyze potential reliability problems during the life cycle of a component, system or equipment, Achieving through this the identification of potential failure modes and their effect on the operation of the system within its advantages is the contribution for decision making, since failures can be anticipated and if they happen they can be solved in quickly, this method increases reliability. (Escalona, Jiménez, & Francisco, 2003)