Design validation - Revision history

S212834: /* Annotated bibliography */

2022-03-23T01:48:43Z

Annotated bibliography

← Older revision		Revision as of 01:48, 23 March 2022
Line 169:		Line 169:
	'''Krüger Nico. (2020). Design Verification vs Design Validation, 6 Tips for Medical Device Makers'''		'''Krüger Nico. (2020). Design Verification vs Design Validation, 6 Tips for Medical Device Makers'''

	This website article is focusing on applying validation and verification in the medical field. However, it ~~gives~~ a good example of how a validation process should be performed, that was used in this article.		This website article is focusing on applying validation and verification in the medical field. However, it provides a good example of how a validation process should be performed, that was used in this article.

	'''Olivier de Weck. 16.842 Fundamentals of Systems Engineering. Lecture 9: Verification and Validation. Fall 2015. Massachusetts Institute of Technology'''		'''Olivier de Weck. 16.842 Fundamentals of Systems Engineering. Lecture 9: Verification and Validation. Fall 2015. Massachusetts Institute of Technology'''

S212834: /* Limitations */

2022-03-23T01:47:23Z

Limitations

← Older revision		Revision as of 01:47, 23 March 2022
Line 149:		Line 149:
	= Limitations =		= Limitations =

	*Validation can give a false idea that a product is good. Indeed, validation needs to be well performed to be efficient. Otherwise, it might create a greater risk for the project and be a waste of time and money. Validation is not only about ticking boxes to check if your product is good or not, but also about asking the right questions. A very important part of the quality of a validation relies on whether the definition of user needs and requirements is adequate or not. If not, then the product is validated according to wrong standards and will not satisfy user needs, while the designers will believe that it does <ref name="MIT"/>. Consequently, the validation must also be validated.		*Validation can give a false idea that a product is good. Indeed, validation needs to be well performed to be efficient. Otherwise, it might create a greater risk for the project and be a waste of time and money. Validation is not only about ticking boxes to check if your product is good or not, but also about asking the right questions. A very important part of the quality of a validation relies on whether the definition of user needs and requirements is adequate or not. If not, then the product is validated according to wrong standards and will not satisfy user needs, while the designers will believe that it does <ref name="MIT"/>. Consequently, the validation itself must also be validated.

	*Validation takes time. It is a long process that must be applied several times during the project. Thus, it can slow down the project a lot, even more so if the validation fails multiple times. Most projects have restricted timelines and managers cannot afford to test every single idea before implementing it. It is then up to the project managers to measure the risk of taking a decision without validating it, but it can sometimes be worth it.		*Validation takes time. It is a long process that must be applied several times during the project. Thus, it can slow down the project a lot, even more so if the validation fails multiple times. Most projects have restricted timelines and managers cannot afford to test every single idea before implementing it. It is then up to the project managers to measure the risk of taking a decision without validating it, but it can sometimes be worth it.

S212834: /* Limitations */

2022-03-23T01:46:35Z

Limitations

← Older revision		Revision as of 01:46, 23 March 2022
Line 153:		Line 153:
	*Validation takes time. It is a long process that must be applied several times during the project. Thus, it can slow down the project a lot, even more so if the validation fails multiple times. Most projects have restricted timelines and managers cannot afford to test every single idea before implementing it. It is then up to the project managers to measure the risk of taking a decision without validating it, but it can sometimes be worth it.		*Validation takes time. It is a long process that must be applied several times during the project. Thus, it can slow down the project a lot, even more so if the validation fails multiple times. Most projects have restricted timelines and managers cannot afford to test every single idea before implementing it. It is then up to the project managers to measure the risk of taking a decision without validating it, but it can sometimes be worth it.

	*Validation deals with user needs. But does the user know exactly what he really needs? Validation is what we can call an evolutionary process. It allows evaluating designs by comparing them to existing ones and it also sticks to user needs. The best that we can hope to get from this is an improved version of an existing product. But it doesn’t really leave room for revolutionary designs, which can sometimes create new needs for the users. The biggest innovations are often revolutionary, such as the development of the internet which became a need through the years, while it ~~couldn’t be~~ before it existed<ref name="Youtube"/>.		*Validation deals with user needs. But does the user know exactly what he really needs? Validation is what we can call an evolutionary process. It allows evaluating designs by comparing them to existing ones and it also sticks to user needs. The best that we can hope to get from this is an improved version of an existing product. But it doesn’t really leave room for revolutionary designs, which can sometimes create new needs for the users. The biggest innovations are often revolutionary, such as the development of the internet which became a need through the years, while it was not before it existed<ref name="Youtube"/>.

	*Designs keep increasing in complexity. This point is a key challenge in various domains. The variables to create the right product are ever-changing and with them, outdated methods of validation fail to produce the right product quite consistently.		*Designs keep increasing in complexity. This point is a key challenge in various domains. The variables to create the right product are ever-changing and with them, outdated methods of validation fail to produce the right product quite consistently.

S212834: /* Limitations */

2022-03-23T01:45:47Z

Limitations

← Older revision		Revision as of 01:45, 23 March 2022
Line 153:		Line 153:
	*Validation takes time. It is a long process that must be applied several times during the project. Thus, it can slow down the project a lot, even more so if the validation fails multiple times. Most projects have restricted timelines and managers cannot afford to test every single idea before implementing it. It is then up to the project managers to measure the risk of taking a decision without validating it, but it can sometimes be worth it.		*Validation takes time. It is a long process that must be applied several times during the project. Thus, it can slow down the project a lot, even more so if the validation fails multiple times. Most projects have restricted timelines and managers cannot afford to test every single idea before implementing it. It is then up to the project managers to measure the risk of taking a decision without validating it, but it can sometimes be worth it.

	*Validation deals with user needs. But does the user know exactly what ~~they~~ really ~~need~~? Validation is what we can ~~all~~ an evolutionary process. It allows evaluating ~~design~~ by comparing them to existing ones and sticks to user needs. The best that we can hope to get from this is an improved version of an existing product. But it doesn’t really leave room for revolutionary designs, which can sometimes create new needs for the users. The biggest innovations are often revolutionary, such as the development of the internet which became a need through the years, while it couldn’t be before it existed<ref name="Youtube"/>.		*Validation deals with user needs. But does the user know exactly what he really needs? Validation is what we can call an evolutionary process. It allows evaluating designs by comparing them to existing ones and it also sticks to user needs. The best that we can hope to get from this is an improved version of an existing product. But it doesn’t really leave room for revolutionary designs, which can sometimes create new needs for the users. The biggest innovations are often revolutionary, such as the development of the internet which became a need through the years, while it couldn’t be before it existed<ref name="Youtube"/>.

	*Designs keep increasing in complexity. This point is a key challenge in various domains. The variables to create the right product are ever-changing and with them, outdated methods of validation fail to produce the right product quite consistently.		*Designs keep increasing in complexity. This point is a key challenge in various domains. The variables to create the right product are ever-changing and with them, outdated methods of validation fail to produce the right product quite consistently.

S212834: /* Limitations */

2022-03-23T01:44:30Z

Limitations

← Older revision		Revision as of 01:44, 23 March 2022
Line 151:		Line 151:
	*Validation can give a false idea that a product is good. Indeed, validation needs to be well performed to be efficient. Otherwise, it might create a greater risk for the project and be a waste of time and money. Validation is not only about ticking boxes to check if your product is good or not, but also about asking the right questions. A very important part of the quality of a validation relies on whether the definition of user needs and requirements is adequate or not. If not, then the product is validated according to wrong standards and will not satisfy user needs, while the designers will believe that it does <ref name="MIT"/>. Consequently, the validation must also be validated.		*Validation can give a false idea that a product is good. Indeed, validation needs to be well performed to be efficient. Otherwise, it might create a greater risk for the project and be a waste of time and money. Validation is not only about ticking boxes to check if your product is good or not, but also about asking the right questions. A very important part of the quality of a validation relies on whether the definition of user needs and requirements is adequate or not. If not, then the product is validated according to wrong standards and will not satisfy user needs, while the designers will believe that it does <ref name="MIT"/>. Consequently, the validation must also be validated.

	*Validation takes time. It is a long process that must be applied several times during the project. Thus, it can slow down the project a lot, even more so if the validation fails multiple times. Most projects have restricted timelines and cannot afford to test every single idea before implementing it. It is then up to the project managers to measure the risk of taking a decision without validating it, but it can sometimes be worth it.		*Validation takes time. It is a long process that must be applied several times during the project. Thus, it can slow down the project a lot, even more so if the validation fails multiple times. Most projects have restricted timelines and managers cannot afford to test every single idea before implementing it. It is then up to the project managers to measure the risk of taking a decision without validating it, but it can sometimes be worth it.

	*Validation deals with user needs. But does the user know exactly what they really need? Validation is what we can all an evolutionary process. It allows evaluating design by comparing them to existing ones and sticks to user needs. The best that we can hope to get from this is an improved version of an existing product. But it doesn’t really leave room for revolutionary designs, which can sometimes create new needs for the users. The biggest innovations are often revolutionary, such as the development of the internet which became a need through the years, while it couldn’t be before it existed<ref name="Youtube"/>.		*Validation deals with user needs. But does the user know exactly what they really need? Validation is what we can all an evolutionary process. It allows evaluating design by comparing them to existing ones and sticks to user needs. The best that we can hope to get from this is an improved version of an existing product. But it doesn’t really leave room for revolutionary designs, which can sometimes create new needs for the users. The biggest innovations are often revolutionary, such as the development of the internet which became a need through the years, while it couldn’t be before it existed<ref name="Youtube"/>.

S212834: /* Example of a validation process */

2022-03-23T01:43:00Z

Example of a validation process

← Older revision		Revision as of 01:43, 23 March 2022
Line 143:		Line 143:
	Let's try to apply the whole validation process to a simple product. A field where validation is particularly important is the medical field where standards are very strict and require a particularly documented validation. The example chosen is a ventilator that allows a patient to continue breathing<ref name="Perforce"/>.		Let's try to apply the whole validation process to a simple product. A field where validation is particularly important is the medical field where standards are very strict and require a particularly documented validation. The example chosen is a ventilator that allows a patient to continue breathing<ref name="Perforce"/>.
	#'''Define the users of the system'''<p>Being a medical device, this ventilator will most certainly be used by nurses or doctors. In preparation for the next step, it should be noted that these are experienced users so some of the ease of use needs may be overlooked.</p>		#'''Define the users of the system'''<p>Being a medical device, this ventilator will most certainly be used by nurses or doctors. In preparation for the next step, it should be noted that these are experienced users so some of the ease of use needs may be overlooked.</p>
	#'''Define the needs of these users'''<p>There are many needs here, so we will select one and analyze it in more detail. Users may need to move a patient in the hospital. So one of the user needs could be simply defined as: "the ventilator can be used ~~when transporting~~ patients in the hospital". We can note that the need specifies that these movements are done in the hospital. Ambulance trips for example would be a second need.</p>		#'''Define the needs of these users'''<p>There are many needs here, so we will select one and analyze it in more detail. Users may need to move a patient in the hospital. So one of the user needs could be simply defined as: "the ventilator can be used during transportation of patients in the hospital". We can note that the need specifies that these movements are done in the hospital. Ambulance trips for example would be a second need.</p>
	#'''Define the associated requirements'''<p>Possible requirements associated with this need would be :</p><p>-The ventilator can be carried by 15 hospital staff members</p><p>-The ventilator continues to operate according to its specifications while being moved</p><p>-The ventilator can pass through doors, corridors, or elevators in a typical hospital</p>		#'''Define the associated requirements'''<p>Possible requirements associated with this need would be :</p><p>-The ventilator can be carried by 15 hospital staff members</p><p>-The ventilator continues to operate according to its specifications while being moved</p><p>-The ventilator can pass through doors, corridors, or elevators in a typical hospital</p>
	#'''Validate these requirements'''<p>Assuming we have a working prototype, we can test all these requirements with a simple usability test, by trying the prototype under the specified conditions and collecting user feedback. </p><p>If we want to add quantified measurements, we can add to these tests time measurements, for example, to be able to evaluate the ease of this transportation. It will then be possible to compare it to other ventilators, or to standards imposing a maximum duration for example.</p>		#'''Validate these requirements'''<p>Assuming we have a working prototype, we can test all these requirements with a simple usability test, by trying the prototype under the specified conditions and collecting user feedback. </p><p>If we want to add quantified measurements, we can add to these tests time measurements, for example, to be able to evaluate the ease of this transportation. It will then be possible to compare it to other ventilators, or to standards imposing a maximum duration for example.</p>

S212834: /* Example of a validation process */

2022-03-23T01:42:20Z

Example of a validation process

← Older revision		Revision as of 01:42, 23 March 2022
Line 143:		Line 143:
	Let's try to apply the whole validation process to a simple product. A field where validation is particularly important is the medical field where standards are very strict and require a particularly documented validation. The example chosen is a ventilator that allows a patient to continue breathing<ref name="Perforce"/>.		Let's try to apply the whole validation process to a simple product. A field where validation is particularly important is the medical field where standards are very strict and require a particularly documented validation. The example chosen is a ventilator that allows a patient to continue breathing<ref name="Perforce"/>.
	#'''Define the users of the system'''<p>Being a medical device, this ventilator will most certainly be used by nurses or doctors. In preparation for the next step, it should be noted that these are experienced users so some of the ease of use needs may be overlooked.</p>		#'''Define the users of the system'''<p>Being a medical device, this ventilator will most certainly be used by nurses or doctors. In preparation for the next step, it should be noted that these are experienced users so some of the ease of use needs may be overlooked.</p>
	#'''Define the needs of these users'''<p>There are many needs here, so we will select one and analyze it in more detail. Users may need to move a patient in the hospital. So the user ~~need is~~ simply defined as : "the ventilator can be used when transporting patients in the hospital". We can note that the need specifies that these movements are done in the hospital. Ambulance trips for example would be a second need.</p>		#'''Define the needs of these users'''<p>There are many needs here, so we will select one and analyze it in more detail. Users may need to move a patient in the hospital. So one of the user needs could be simply defined as: "the ventilator can be used when transporting patients in the hospital". We can note that the need specifies that these movements are done in the hospital. Ambulance trips for example would be a second need.</p>
	#'''Define the associated requirements'''<p>Possible requirements associated with this need would be :</p><p>-The ventilator can be carried by 15 hospital staff members</p><p>-The ventilator continues to operate according to its specifications while being moved</p><p>-The ventilator can pass through doors, corridors, or elevators in a typical hospital</p>		#'''Define the associated requirements'''<p>Possible requirements associated with this need would be :</p><p>-The ventilator can be carried by 15 hospital staff members</p><p>-The ventilator continues to operate according to its specifications while being moved</p><p>-The ventilator can pass through doors, corridors, or elevators in a typical hospital</p>
	#'''Validate these requirements'''<p>Assuming we have a working prototype, we can test all these requirements with a simple usability test, by trying the prototype under the specified conditions and collecting user feedback. </p><p>If we want to add quantified measurements, we can add to these tests time measurements, for example, to be able to evaluate the ease of this transportation. It will then be possible to compare it to other ventilators, or to standards imposing a maximum duration for example.</p>		#'''Validate these requirements'''<p>Assuming we have a working prototype, we can test all these requirements with a simple usability test, by trying the prototype under the specified conditions and collecting user feedback. </p><p>If we want to add quantified measurements, we can add to these tests time measurements, for example, to be able to evaluate the ease of this transportation. It will then be possible to compare it to other ventilators, or to standards imposing a maximum duration for example.</p>

S212834: /* A/B testing */

2022-03-23T01:40:52Z

A/B testing

← Older revision		Revision as of 01:40, 23 March 2022
Line 136:		Line 136:
	#'''Collecting the results of your test'''		#'''Collecting the results of your test'''
	#'''Changing or not the design according to which version is best'''		#'''Changing or not the design according to which version is best'''

			This method is a very good way to validate and justify through numbers any design choice of your product. However, it is again a very long process and it is almost impossible to apply it to every single design choice. It should be used mostly for decisions evaluated as critical.

	== Example of a validation process ==		== Example of a validation process ==

S212834: /* A/B testing */

2022-03-23T01:38:53Z

A/B testing

← Older revision		Revision as of 01:38, 23 March 2022
Line 126:		Line 126:
	==== A/B testing ====		==== A/B testing ====

	A/B testing is a technique that consists of proposing several variants of the same object that differ according to a single criterion in order to determine which version gives the best results with consumers. It is a technique particularly used in online communication where it is possible to test with a sample of people several versions of the same web page, the same mobile application, the same email, or an advertising banner in order to choose the one that is the most effective and to use it on a large scale.		A/B testing is a technique that consists of proposing several variants of the same object that differ according to a single criterion in order to determine which version gives the best results with consumers. In fact, it is also possible to modify several criteria at once to speed up the process, or to propose more than 2 different versions for a single criterion. This technique is particularly used in online communication where it is possible to test with a sample of people several versions of the same web page, the same mobile application, the same email, or an advertising banner in order to choose the one that is the most effective and to use it on a large scale.

	For example, if we apply A/B testing to the design of a car, it will consist of varying the color of the car, its maximum speed, its shape, or perhaps an element of its packaging and observing which version will be the most convincing <ref name="AB"/>. The process of A/B testing is therefore:		For example, if we apply A/B testing to the design of a car, it will consist of varying the color of the car, its maximum speed, its shape, or perhaps an element of its packaging and observing which version will be the most convincing <ref name="AB"/>. The process of A/B testing is therefore:

	#'''Defining your objective'''		#'''Defining your objective''' (what you want to improve about your design)
	#'''Identifying your target precisely'''		#'''Identifying your target precisely''' (what part of your design will be modified)
	#'''Varying an element of your product'''		#'''Varying an element of your product'''
	#'''Letting your audience accept these variations, interact and react to your new proposals'''		#'''Letting your audience accept these variations, interact and react to your new proposals'''
	#'''Collecting the results of your test'''		#'''Collecting the results of your test'''
	#'''~~Adapting your product~~ design '''~~(but it could also be your communication, your sales strategy, etc.)~~		#'''Changing or not the design according to which version is best'''

	== Example of a validation process ==		== Example of a validation process ==

S212834: /* A/B testing */

2022-03-23T01:27:51Z

A/B testing

← Older revision		Revision as of 01:27, 23 March 2022
Line 126:		Line 126:
	==== A/B testing ====		==== A/B testing ====

	A/B testing is a technique that consists of proposing several variants of the same object that differ according to a single criterion in order to determine which version gives the best results with consumers. It is a technique particularly used in online communication where it is ~~now~~ possible to test with a sample of people several versions of the same web page, the same mobile application, the same email, or an advertising banner in order to choose the one that is the most effective and to use it on a large scale.		A/B testing is a technique that consists of proposing several variants of the same object that differ according to a single criterion in order to determine which version gives the best results with consumers. It is a technique particularly used in online communication where it is possible to test with a sample of people several versions of the same web page, the same mobile application, the same email, or an advertising banner in order to choose the one that is the most effective and to use it on a large scale.

	For example, if we apply A/B testing to the design of a car, it will consist of varying the color of the car, its maximum speed, its shape, or perhaps an element of its packaging and observing which version will be the most convincing <ref name="AB"/>. The process of A/B testing is therefore:		For example, if we apply A/B testing to the design of a car, it will consist of varying the color of the car, its maximum speed, its shape, or perhaps an element of its packaging and observing which version will be the most convincing <ref name="AB"/>. The process of A/B testing is therefore: