What is validation?
Validation is a central activity in data processing. The validation of data is therefore part of the software development process as well as during software operation, i.e. during the actual use of software programs.
What does validation mean?
Validation in computer science means
“to determine the importance, validity, value of something”
In general, validation can also be understood as a test that basically checks whether a software product is suitable for its intended use or meets the desired requirements. The expectations of the technical experts and users are also incorporated into such evaluations.
What does validation mean? Quality check – Confirm validity and therefore correctness
Practical example VDI Guideline 2206
One area of application is the VDI Guideline 2206 which defines the development of cyber-physical mechatronic systems . This guideline covers the entire development process and the associated individual steps. The aim of the quality guideline is to form a logical basis for the development of Certified Professional for Medical Software (CPMS).
What exactly do we validate?
In software development, we validate requirements and also when using computer programs, we validate variables and the values and data stored in these program variables. It is equally important to validate the processes.
Why validate data?
Validation methods
Within a program, a validation method defines a validity check, whereby data can be validated in a meaningful way, for example with regard to:
- Value ranges, i.e. the value of a variable lies within a defined value range
The value range can be used to determine for a measurement method within software, for example, whether the test results recorded with the data values actually capture what is to be determined.
Validate data to avoid error messages because the entered data has not been validated
Optimization of validation processes
The digital processes for verification are of central importance for the validation of data in real time.
Optimize products, software or services with validated data
Anyone who masters this technology can use validated data to optimize products, software or services.
The importance of processes in validation and how to optimize them to be more effective and efficient:
Processes are an important part of validation as they ensure that all steps in the validation process are clearly defined and documented.
By using processes, companies can make validation more effective and efficient by ensuring the repeatability and traceability of processes, eliminating unnecessary steps and identifying and removing bottlenecks.
Process optimization can also help to save time and resources and improve validation results.
How verification techniques such as tests and inspections can be used to validate products or services:
Verification techniques such as tests and inspections can be used specifically to validate products or services to ensure that they meet the requirements.
For example, tests can be carried out to check the functionality and performance of products or services.
Inspections can be used to check conformity with technical specifications and standards. The choice of verification technique depends on the type of product or service to be validated and the requirements.
How verification and validation affect the validation process
Validation refers to checking the conformity of products or services with the requirements, while verification refers to checking the correctness of products or services.
Validations refer to the validation process, including the checking of products or services for conformity with requirements, while verifications refer to the performance of tests and inspections to check the conformity of products or services with requirements.
All steps in the validation process should be clearly defined and documented to avoid misunderstandings.
How does the validation process work?
The validation process is a systematic approach to checking the conformity of products or services with the requirements. The validation process usually involves several steps, which can be described as follows:
- Definition of the requirements: The validation process begins by defining the requirements for the product or service to be validated. This may include, for example, the review of technical specifications, standards, customer requirements and regulatory requirements.
- Planning: The next step is to create a validation plan that defines the objectives, scope, timeframe, responsibilities, resources and verification and validation techniques for the validation process. This plan serves as a roadmap for the validation process.
- Carrying out verifications and validations: The validation process involves carrying out verifications and validations to ensure that the product or service meets the requirements. Verifications include, for example, tests, inspections and analyses to ensure that the product or service is correct and complete. Validations include, for example, checking whether the product or service meets the requirements and fulfills the intended purpose.
- Reporting: Once the validation process is complete, a report is produced summarizing the results of the validation process. The report usually contains information on the verifications and validations carried out, the results, the deviations and the corrective measures.
- Follow-up: After the validation process has been completed, follow-up measures may be necessary to ensure that the product or service continues to meet the requirements. This can include, for example, monitoring trends, conducting reviews or carrying out revalidations.
Best practices to ensure successful validation
Some best practices for using verification and validation in the singular and plural are:
- Clearly define what verification and validation mean in context
- Ensure that all steps in the validation process are clearly defined and documented
- Check whether the verification and validation processes meet the requirements
- Involvement of experts and stakeholders in the validation process
- Continuous monitoring of data using automated tests and monitoring
Why is data validation so important in software development?
Validation is therefore of central importance in the qualitative improvement of software .
Data validation is therefore essential during the development and use of software.
Validate data from the first prototype to the current software release
According to this understanding, validation must be continuous over the entire development period in all software releases.
software releases
and with the start of the first
prototypes
from the beginning.
This is the only way to ensure that the software can later be used without errors.
Potential of data validation to increase software quality
The potential and positive effects of early detection of problems and errors are correspondingly high. However, due to time and cost constraints, which practically all software development processes are subject to, it is not possible to validate as extensively as desired. The aim must therefore be effective and efficient validation based on an understanding of the system and model. Unfortunately, from today’s perspective, validation is neither comprehensively formalized nor universally valid and consistently methodologically supported. Rather, validation is essentially experience-based and characterized by the individual skills of the software developers as well as discipline-, task- and application-specific.
With methodology and standardized analysis for established data analysis
By applying methodical principles as well as continuous and standardized documentation and analysis, routines and models can be reused and specific reference processes can be derived that can be used for future software developments.
Practical examples of data validation
Validation provides a statement as to whether a created model adequately describes a real system and thus also fulfills any unspecified requirements.
Validation: Are specified requirements fulfilled?
To illustrate this, here is a current example of problems that have occurred in the past with the air conditioning systems of the ICE fleet: At very high temperatures, the air conditioning systems of the ICE3 trains repeatedly fail. Deutsche Bahn ultimately confirmed to the Federal Railway Authority that the cooling system for the older ICE3 fleet was only specified for outside temperatures of up to 32 °C.
The specified requirement – cooling up to an outside temperature of 32 °C – was met and certainly verified. However, it appears that it has not been conclusively determined whether this requirement also corresponds to customer needs over the useful life, i.e. whether it is valid.
Initiate feature requests and optimization based on validated data
As a consequence and countermeasure, today’s systems are designed for outside temperatures in the range of 40 °C and are therefore better suited to the changes in climatic conditions over the past 20 years. In industrial practice, the term “safeguarding” is often used. This summarizes all test activities of a development project, from material testing to test drives or endurance tests.
Air conditioning systems are also software-controlled components, so the example provides a practical and easy-to-understand example of the central importance of validation.
With the
clustering algorithm
you will find an algorithm-based and concrete solution on how to validate data.
