>> Home >> Requirements Process and Management Services
We offer a range of requirements management and systems engineering support services as well as configuration management, risk management, and PRINCE based project management consultancy. Our requirements services form the foundation of our offerings, often run as a project support office, and include :
This page also provides a basic introduction to requirements management process concepts, which are summarised in our draft Requirements Process Primer document which is available as a free download.
The adoption of a requirements tool or a process model is not a golden bullet. They will not necessarily provide a quick fix in terms of immediately stopping problems from occurring or bring immediate increases in productivity. Often the reverse is true - the initial result is to bring out many of the existing problems, you still feel the pain, but you'll have more time to effectively manage the problems, but over time you'll be able to correct the underlying problems. With process feedback and implementation outcomes such as buy-in, training, and cultural changes, you'll see benefits, and you can start to tune the process and improve it to fit your needs, and start on the continuous improvement cycle using the Plan-Do-Check-Act cycle. Beaver has experience in all these areas and will be happy to discuss how our experience can help you.
A Simple Requirements Management Process Model / Systems Engineering Process Model Concepts
This section outlines a general lifecycle process model and the factors to be considered when setting up requirements tools. We have summarised a number of the process elements for public use. This part of the process focuses on the requirements flow. Sub-models in our process deal with elication and ensuring completeness of each specification using standard contents templates to prompt for information to be added or explicitly add a not used comment. Our requirements and V&V process and the database / information model concepts have been developed through practical application on a range of projects. They have been reported at conferences as detailed in our news items for June 1999 and September 2000. See our case studies section for details of the how this expertise has been used when working with our clients. Our web links page lists a number of other requirements management sites. To see an example of DOORS based traceability and the creation of a simple traceability matrix see our paper on Joyce Ludwigs Requirements site.
A simple process model is shown above. Placing your cursor over each process block will provide you a summary of the inputs, transformation activities, and outputs. Process boxes with a red outlined blocked can be double clicked to open a more detailed page, giving further details of the inputs, outputs, and activities, and where appropriate we have included sample check lists, and screen shots of DOORS configured to support a work flow based process. A list of documents and the stage at which they are produced is listed here <to be added>.The above process model looks complex because of the concurrency and feedback flows that have been included. You'll also need to be consider how this process spreads out over time. In a moment we'll look at several views of the process and supporting information, some with the help of animation. You'll see that when the process is stretched out against time it is not as difficult or complex as it at first appears, and with our step by step guides you'll soon be in control.
|Our draft white paper is available here which provides a more detailed overview of the process, sample information model detailing attributes, views, and reports, and how the process model, specification tree, and information model fit into change control, risk management, and iterative based life cycle models. The paper examines the design of the information model for requirements and test traceability. Such traceability forms the basic inputs for a number of process activities. Note this is a draft and is provided as-is. Please see the conditions of use on the front of the document.|
You can review a Powerpoint animation online here <to be added> which will step you through the major process steps against time or you can download the Powerpoint file here <to be added>. When viewing the animation remember this a simple representation - consider the additional complexity when you consider individual requirements progress changes, rework and test failures requiring requirement, test spec, or test harness changes. It is this level of concurrency that causes problems if not adequately managed.
As an alternative perspective you might wish to review the "V" model based specification tree and its animated version. The specification tree Powerpoint files are also available for download here <to be added>. The process model produces the documents shown in the specifications tree and conducts appropriate quality tests, trade-offs, and results reviews. The process model produces the documents in the specification tree while the information model is the basis on which the underlying data is managed, link, stored, viewed, and operated on.
A process model and specification tree are useful to (a) understand the needed integrity of the model and enforce change control and configuration management of the set of documents which exist at a point in time, and (b) acts as the basis to generate metrics to indicate both the progress through process (e.g. number of client requirements, number of accepted requirements, number allocated to technical solution, number tested) and the quality of the products produced using data from the various reviews (number of ambiguous requirements, problems found), issues and assumptions associated with a requirement or product design. Beaver believes it is important not only to measure pure progress but the quality. Progress is OK only if an acceptable quality level if being achieved otherwise rework from problems will arise. In addition quality provides feedback to drive process improvement initiatives such as those included as part of the SEI's CMMI.
The process model, even a simple one such as above, are not the information model. Informational models are best considered as relational database structures. A simple information model is shown here <to be added>. Each file or module has attributes, which are selectively shown as a set of views, or reports. The information models seeks to reuse information rather than having multiple copies, and show the relationships between the various files or modules. Additional code or functions are added to applications running the models to ensure integrity e.g. if a parent requirement is changed then links to child requirements or test results are shown as suspect as they can no longer be assured to meet the parent because it has changed. A manual review is needed.
Own staff not available, no in-house capability, need a fresh view? Do you need to create a specification, improve an existing one, or apply other requirements management processes? Whether you own a dedicated tool, or you normally use paper and MS Office, or you just need the work done we can help you immediately. For example;
Having a defined process is an basic building block in developing a strong system engineering or project management capability and is also important when demonstrating compliance to the process orientated ISO 9001 : 2000 and when benchmarking to the SEI's Capability Maturity Model Integrated (CMMI) or the OGC's Project Management Maturity Model (PMMM). We can support your process management activites using Six-Sigma based methods such as
We also provide training and mentoring in requirements management concepts and in training and mentoring users in using tools such as DOORS to implement requirements management through the life cycle. Our approach is to use a modular approach to training. The outcome of each training module is to provide sufficient DOORS functionality to support specific requirements management and systems engineering tasks. Our approach reflects our prefered way of using DOORS, which is to setup and use specific views and key funcitions, wizards and addins, which relate to process steps in your management system.
Rather than provide a 2-3 day course attendees only need to learn whats necessary before going back to the job and whats needed. For example most users do not need administrator or dxl skills, many just need to find particulary views in the tool, fill in several attributes, or raise or approve changes. Others just need to track progress. Attend whats needed and then put the learnt skills into practice doing assigned work. We can provide on-site mentoring or phone or email support post training. Having mastered the tasks from one module attendees can then take the next module. We have used this approach with a number of clients, often combined with some tool configuration and process management work. Ofcourse it is possible to run a combination of the modules as a 2 day course which works so long as attendees can go back to performing specific tasks quickly as early use reinforces whats been learnt and aids retention. Training sessions can be based round the following modules :
Beaver develops and advises clients in the IT infrastructures to implement requirements and systems engineering processes, integrating and interfacing project management (process, tasking allocation, planning, risk management), problem and defect reporting, consolidated change change systems, with requirements tools and analysis and design tools. These can be developed as single PC systems, integrating various applications, run as multi-user versions running over networks, or implement using intranet or internet technologies, giving access to requirements, change management, and project management support using an internet browser with backend systems linking back to the core applications.To set up a requirements tool such as DOORS and customise it to fit your process we undertake the following activites:
Project fast-track service - gets you up and running quickly. We review your existing process and documents for a project, understand the relationships between each document, and develop an information model. We do by working with you using a mixture of offline information gathering followed up by a facilitated joint development workshop with the users or potential users of DOORS or other requirements tool.
Customise requirements tools to make them user friendly - often customising menus so they reflect the process steps. This allows easier user interfaces and means users do not need to learn all the intricate tool commands. They simply drive a process based menu which presents them with options for showing various process steps, reviews, reports, metrics and integrity tool. This allows them to focus on doing requirements, systems and development work rather than using the tool. It reduces learning time, re-enforces process steps, and can reduce errors by including code to identify suspect integrity errors and certain types of requirements problems.
Writing DXL tools to assist in importing specifications and other documents, flagging likely requirements and also those that are likely to be problematic. You can use the same tools and re-applying them as each level of specification is produced. These tools do not automate the process or eliminate people from the process, far from it, but rather they are used to rather kick start the process, apply a level of consistency, and allow your personnel to focus on the real issues. In our experience many organisations do not even perform these tasks manually so the tools help to introduce a process for best practice.
Developing DXL based metrics or statistical tools to pull information out from the data and show the progress through the process, the quality of the products produced by the process, and the quality of the process its self, for example defaults captured or missed at each stage or to show the stability or changes in requirements, design, or testing and the associated unclosed threats, risks, issues and assumptions which should diminish over time.
Customise menus and provide productivity tools using DOORS DXL or C/C++ and VBA to interface to the API's for other tools, and to integrate sets of tools using VBA and technologies such as Microsoft Digital Dashboard bringing together MS Outlook, Project, and third party tools. Beaver believes productivity can be improved by doing work in the tools themselves rather than using them to just record the abstract results.
Develop custom bridges between DOORS and other tools, that have MS-SQL, My-SQL or Access repositorys using DOORS DXL, VBA, Perl, or PHP.
Integrate DOORS with design and analysis, configuration management, defect tracking, project management, risk management, and safety and reliability analysis tools.
Development information models, attributes, views, and reports to implement defined processes and reflect suitable document structures and be capable of producing the required documents. Beavers approach is to provide traceability between the process and the set up of information model, tools, reports, and report content, and guidelines for other activities such as technical reviews, design reviews, and peer reviews. The process model itself can be traced to standards and CMM's as well. We provide process based information models / requirements databases in DOORS and Requisite 'Pro formats. Please contact us for full details or see here for an outline preview.
Process definition, developing intranet and paper based manuals and interfaces to tools implementing parts of the process. Beaver has experience including standards such as IEEE 1220, EIA 632, MIL STD 498, MIL STD 2167A, public methods such as RUP, MSF, DSDM, SSADM, and RtP for example. We have particular experience in auditing existing processes and effectiveness against capability maturity models e.g. CMMI. We also have experience in processes for specialist disciples such as safety, RAM, and ILS working to IEC 61508, DEF STAN 00-41, 00-55, 00-56, 00-60.
Beaver is experienced in analysing processes and how they are implemented to identify potentials errors and to develop control measures (e.g. check list items, integrity checkers, changed processes or procedures, training, monitoring), based round process analysis practices in DEF STAN 00-55 and ISO 9000 : 2000. We under take BPR studies, identifying duplicated tasks, data or information re-entry in multiple areas causing duplication causing inappropriate resource allocation and difficulties with configuration management as data becomes varied, uncontrolled and out of date. We look at the time it takes to complete various activities, such as document production or testing, the level of rework and the causal problems. The information gained is used to redefined the processes used to reduce cycle time, reduce resource requirements, improve quality and reduce rework, and providing a more motivation process and environment.
Extending traditional information models, with extensive traceability, such as holding functional and functional interface requirements and product and architectural solutions to be able to perform HAZOP and FMEA's, record derived requirements and actions, so formally capturing the design and its analysis, leading to derived requirements. Because the work has been directly captured in DOORS for example as opposed to merely entering data as unrelated entries or not at all, the exact status of the design is visible and traceable, and as the process for example has generated the safety requirements, DOORS can be used to add the resulting argument and supporting evidence, so the safety case can be easily constructed. Failures and hazards can also be exported to other analysis tools e.g. fault tree and simulation programmes and other results entered back in.
Implement information model structures and reporting capabilities which develop specifications and reports developed from a number of files or modules. A common problem Beaver has encountered is when projects develop the information models to reflect modules and desired documents on a one for one basis. Most specification formats can not be achieved using a single file or module. They are best developed by exporting different sets of information from various modules. Beaver has experience in developing and using a variety of reporters and now uses tools such as those developed by Telelogic DocExpess and Rational SoDA and can develop templates, views, and customise DXL or code to extract and format data as need to develop a variety of documents such as specifications, change analysis proforma's, test specifications, test specification results, FMEA and HAZOP reports, and risk and issues registers.
Return to the requirements information page