Model-based requirements management methods and tools-Pitaya Software

Model-based requirements management methods and tools

Zu Tao （Pitaya System Engineering）

Views

2019-5-27

Preface

Requirements management is an engineering problem that everyone is concerned about, and it seems to be untechnical, but it is difficult to solve. This paper proposes a model-based requirements management method, and based on this theory, the mainstream modeling tool EA is selected, and the corresponding requirements management tool is customized. It is hoped that peers who are facing the same requirements management challenges will be inspired.

Body

Requirements are the basis of a lot of work:

The scope management and workload estimation of the project should be based on requirements

A sufficient analysis of requirements, such as functionality, extensions, and performance, is required to effectively design the framework and operation process of the software.

Testing also requires the design of test cases based on requirements and the development of verification standards.

A lot of the work of a product manager is based on requirements, such as product requirements analysis, product design, and product release management.

Demand management is also the most problematic link in the software development process and system engineering, whether it is a control system on aviation, aerospace and automobiles with high reliability requirements, or a business system for enterprise informatization, the quality of requirements has always been paid attention to, but has never been effectively solved.

The working process of requirements is generally as follows:

First, the original request of the user is received, and then the demand personnel analyze the user's requirements and describe the use case clearly.

On the basis of clear user needs, further product demand analysis is carried out to define the functional modules of the system;

The developer and the user sign a demand contract to clarify what to develop, how much work to invest, estimate the cost and time, and determine the budget.

The development team then develops based on the requirements, and in the process, the user's requirements are subject to change control so that the quality product can be delivered on the expected time.

When the development is completed, it is verified based on the requirements, and only those that are qualified can become the products delivered to users.

Although the development process can be carried out iteratively, from the user's point of view, a high-level development team should confirm the requirements before development, and the delivered product should also meet the quality requirements.

Based on the above requirements workflow, the following requirements management content can be sorted out:

The user's original request

Models for requirements analysis, e.g., use case models, interface models, and models that describe non-functional requirements.

The system or software itself that is being developed or delivered.

Use cases for requirements validation.

Defects or inconsistencies in requirements found

The management content of requirements seems to be varied, and it should be related to the subsequent design, development, and testing of the artifacts, so that there are more related contents, how to manage them well? The management content of various jobs can be organized into three categories:

Documents, such as requirements documents, design documents, development documents, and test documents.

Item data, e.g. a list of required use cases, a list of designed modules, a directory of developed codes, and a list of tested bugs.

Models, such as requirements models, architecture models, data models, code models, and test models.

Although each content is different, the same form of management method is basically the same. Among the three forms, the model is suitable as a form of specialized analysis and design work; Documentation can be read by everyone, especially those who are not designers and developers; Item data is suitable for tracking management.

Through the above analysis, there is no doubt that the model is the key to the whole research and development process, if there is no adequate analysis and design, the quality of the management content is not good, the organization is not clear, and the best management process and tools are not well managed. It is for this reason that we propose a model-based approach to requirements management, and the main ideas are:

Professional requirements analysis, system design, software design and test analysis design based on the model.

Based on the hierarchical relationship in the logical relationship of the model, the model is turned into a list of items for tracking management, and the focus is to establish the association between various model elements, such as: the implementation relationship between the use cases in the requirements and the components in the design, and the tracking relationship between the bugs in the test and the corresponding requirements use cases or components.

Various documents can be generated based on the model, so that the documents are naturally associated with the model, and can also be managed in the form of a list. By converting the model into list data, it is easy to support the forms of object lists, impact analysis matrices, tracking trees, kanbans, statistical charts, etc., and the requirement difference analysis of each version is also much more convenient.

To this end, we have specially developed a model-based requirements management tool iSpace, which is based on the model generation of the mainstream modeling tool EA for management of itemized data, and can be extended to the models of other modeling tools according to the needs of users in the future.
The following are screenshots and descriptions of some usage scenarios of iSpace.

The elements in the EA model can be easily tracked and managed by using the list view of iSapce.

Based on the full-cycle model of systems engineering or software engineering: requirements model, design model, code model, and test model, all of which can be browsed and managed in the view of the item list. This makes it easy to manage completely, and all kinds of management (requirements management, design management, code management, test management) have a set of methods, some of which overlap, and if they are managed separately in different tools, the correlation between them is a complex work. If all the models are built in one modeling tool, and then correspond to various management views based on the models, the seemingly complex management will be simple together, and many association problems will be easily solved.

As follows, if you edit the corresponding elements and attributes in iSpace, they will be automatically synchronized to the modeling tool and vice versa.

It is also very convenient to select an element, and quickly select other associated elements through the association selector, which is much more convenient to operate, and the key is that the relationship between the corresponding elements on the model diagram can also be automatically synchronized in the modeling tool EA.

In order to compensate for the problem of a relatively simple description of the modeling tool (mainly text-based), a rich text editing area is provided where users can:

Create a table

Paste the image

Upload the video

Hook up documents

This is useful for some work scenarios, such as:

The demand person should put an interface prototype diagram or a professional algorithm diagram under a use case.

The tester finds a bug, takes a screenshot, and puts the screenshot under a bug model element.

Based on the relationship between the elements in the model, you can quickly generate an association context diagram, which is of great value for the impact analysis of requirements changes or bugs.

The following is a screenshot of the working scenarios for various management requirements.

The following is the change impact analysis, after establishing various models (requirement model, design model, code model, test model), if the user proposes a requirement change, you can generate an association context diagram, select a change, and you can view the content of all models affected.

The following is the bug impact analysis, after establishing various models (requirement model, design model, code model, test model), if a bug is found, you can generate an association context diagram, select a bug, and you can view the content of all models affected.

Requirements also need to track various states (proposed, to be developed, to be tested, to be released, published), at present, Kanban is a very practical tracking method, iSpace can automatically assign to the status column of Kanban according to the status of various elements, and can also change the state of elements by dragging. If it is equipped with a large screen, it is a good electronic signage.

Requirement management needs to carry out a variety of statistics, such as demand completion statistics, requirements verification coverage statistics, demand quality statistics, for which a variety of statistical charts are provided, statistical data are from the model, users can convert to the statistical view at any time, real-time understanding of various information of requirements, and guide related development, testing and management work.