US20220066744A1

US20220066744A1 - Automatic Derivation Of Software Engineering Artifact Attributes From Product Or Service Development Concepts

Info

Publication number: US20220066744A1
Application number: US17/409,352
Authority: US
Inventors: Christian Körner
Original assignee: Siemens AG
Current assignee: Siemens AG
Priority date: 2020-08-25
Filing date: 2021-08-23
Publication date: 2022-03-03
Also published as: US11789702B2; US20220066742A1; US20220066743A1; US11586423B2; EP3961382A1; EP3961384A1; EP3961383A1

Abstract

Various embodiments of the teachings herein include a computer-implemented method for automatic derivation of attributes of software engineering artifacts arising from technical boundary condition of products or services comprising: deducing technical requirements based on classifications of the technical boundary conditions; mapping the deduced technical requirements of the artifacts to engineering disciplines and concerns; mapping the calculated engineering artifacts to responsibilities; adapting the classification of the technical boundary conditions based on the evaluation results in iterations; calculating a distribution of the classification space based on a calculation of a multi-selection in technical boundary taxa; and calculating distribution and quartiles.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims priority to EP Application No. 20192815.7 filed Aug. 26, 2020 and EP Application No. 20192519.5 filed Aug. 25, 2020, the contents of which are hereby incorporated by reference in their entirety.

TECHNICAL FIELD

The present disclosure is related to software engineering. Various embodiments of the teachings herein include computer-implemented methods and/or computer systems for automatic derivation of attributes of software engineering artifacts.

BACKGROUND

To achieve a certain product or service quality, artifact quality must be aligned of the variety of engineering disciplines, especially software engineering disciplines depending on the underlying task or challenge. Unaligned artifact quality creates quality deficits. Up to now, no technical solution is provided to this sort of problem which would allow for differentiation between segments of an initial concept or even for prioritization. Organizational or non-technical procedures are applied sometimes to create consistent attributes for software engineering artifacts, meaning engineering goals, but usually fail to create alignment especially for fairly complex systems or frequent changes in the development task.

SUMMARY

In view of this, the teachings of the present disclosure include methods for successful implementations of projects, product, or service developments that allows for an efficient evaluation of an increased number of requirements within less or the same time to calculate a consistent set of software engineering artifact attributes. For example, some embodiments of the teachings herein include a computer-implemented method for automatic derivation of attributes of software engineering artifacts, which attributes arise from technical boundary condition of products or services, comprising the measures: deduction of technical requirements by an automated software-based process based on classifications of the technical boundary conditions, mapping the deduced technical requirements of the artifacts to engineering disciplines and concerns by an automated software-based process, mapping the calculated engineering artifacts to responsibilities, adaption of the classification of the technical boundary conditions based on the evaluation results in iterations, distribution calculation (DC) of the classification space, wherein the distribution calculation of the classification space is based on a calculation of a multi-selection in technical boundary taxa, and a calculation of distribution and quartiles.
In some embodiments, the mapping of the deduced technical requirements of the artifacts to engineering disciplines and concerns at first is kept between a segment choice of the technical requirements and at second for each multi-selection possibility per segment, the relations to all engineering disciplines and concerns are counted.
In some embodiments, the method further comprises the measures: evaluation of a specific multi-selection, wherein the relationships, built in the mapping step, are counted and related to the distribution of possible multi-selections split up into quartiles, and selection of a quartile based on technical requirements, wherein the quartile definition is based on a calculated distribution which entails an algorithm considering all selection possibilities of the classification space.
In some embodiments, the method further comprises evaluation of the mapping results based on software metrics, wherein the software metrics measure the completeness selectivity of the mapping.
In some embodiments, the distribution calculation of the classification space is based on a calculation of weighted multi-selection in technical boundary taxa and a calculation of distribution and quartiles.
In some embodiments, the method includes placing a weight is placed on the selection of a technical boundary condition based on technical requirements, wherein the weight is multiplied with the weight of the combination created by the multi-selection and used the resulting value to look up the quartile in the distribution.
In some embodiments, the method includes placing a weight on the relation between the engineering disciplines and concerns and the products or services development requirements, wherein the relation weight is used as a count absorbed by the weight calculation.
In some embodiments, the distribution calculation results are stored and evaluated for further subjecting the calculation results to a metric based ranking.
As another example, some embodiments include a computer system for automatic derivation of software engineering artifacts, comprising a first subsystem with the components: a classifier software component for the classification of the technical boundary conditions, a calculation software component for the deduction of the technical requirements, a first mapping software component for mapping the technical requirements to engineering artifacts, an I/O-component for receiving the technical boundary conditions data and for providing the calculation results, a storage component, and a second subsystem which provides a distribution calculator software component for the distribution of the classifications, wherein the distribution calculator software component for the distribution of the classifications provides a calculation of multi-selection and a calculation of distribution and quartiles.
In some embodiments, the distribution calculator software component for the distribution of the classifications provides a calculation of weighted multi-selection combinations and a calculation of distribution and quartiles.
In some embodiments, the storage component comprises at least a data base containing relevant data for the mapping processes.
In some embodiments, the distribution calculator software component has access to the storage component for storing the calculation results.
In some embodiments, the first subsystem comprises an evaluation software component for subjecting the calculation results to a metric based ranking.
In some embodiments, the program is executed by a computer, cause the computer to carry out the steps of the methods described herein.
As another example, some embodiments include a provision apparatus for the computer program products described herein, wherein the provision apparatus stores and/or provides the computer program product.

BRIEF DESCRIPTION OF THE DRAWINGS

The accompanying drawings are intended to provide a better understanding of the embodiments described herein. They show embodiments and serve in conjunction with the description to explain principles and concepts of the disclosed subject matter. Other embodiments and many of the cited advantages will emerge in respect of the drawings. The elements of the drawings are not necessarily shown in scale with one another. Identical reference characters here designate identical components or components of comparable effect. The properties, features and advantages of the teachings described herein and the manner in which they are achieved will become clearer and more distinctly comprehensible in conjunction with the description of the exemplary embodiments that follows, these being explained in more detail in conjunction with the FIGS. 1 to 4, in a schematic depiction:

FIG. 1 shows a diagram for an exemplary embodiment of the system with first and second subsystem;

FIG. 2 shows the first subsystem with its components;

FIG. 3 shows the relations between the method steps; and

FIG. 4 shows a diagram for the distribution calculation interaction with modules of the first subsystem.

DETAILED DESCRIPTION

The teachings of the present embodiments may be implemented in a computer-implemented method for automatic derivation of attributes of software engineering artifacts, which attributes arise from technical boundary condition of products or services, comprising the measures: deduction of technical requirements by an automated software-based process based on classifications of the technical boundary conditions, mapping the deduced technical requirements of the artifacts to engineering disciplines and concerns by an automated software-based process, mapping the calculated engineering artifacts to responsibilities, adaption of the classification of the technical boundary conditions based on the evaluation results in iterations, distribution calculation of the classification space, wherein the distribution calculation of the classification space is based on a calculation of a multi-selection in technical boundary taxa and a calculation of distribution and quartiles.
In some embodiments, the multi-selection step provides a sufficiently fine-grained rating of products or services developments and their requirements. While single selection up to now only provide differentiations of about 2·10⁶possibilities, the extension to multi-selection allows for some orders of magnitude more possibilities of about 2·10¹². These numbers are based on an exemplary embodiment of the method, which has 9 taxa, classification dimensions, and therefore a 9-dimensional selection space, with 3 to 16 different selections. The number of possible selections meaning the size of the selection space. Single selection is the product of the numbers of different selections in each taxa. Multiple selection is the product of pow(2,n) of each taxa with n=number of different selections and prioritized multiple selection is the product of pow(c,n) of each taxa with c=number of applicable priorities. Reasonable numbers of c are between 3 and 7.
In some embodiments, the mapping of the deduced technical requirements of the artifacts to engineering disciplines and concerns at first is kept between a segment choice of the technical requirements and at second for each multi-selection possibility per segment, the relations to all engineering disciplines and concerns are counted. This way, the engineering disciplines and concerns are related to the products or services development requirements in a comprehensible way.
In some embodiments, the method further comprises the measures: evaluation of a specific multi-selection, wherein the relationships, built in the mapping step, are counted and related to the distribution of possible multi-selections split up into quartiles, and selection of a quartile based on technical requirements, wherein the quartile definition is based on a calculated distribution which entails an algorithm considering all selection possibilities of the classification space. This allows to relate a given multi-selection to its engineering difficulty. The actual advantage is that the mapping stays constant in time and effort irrespective of the size of the mapping space. The prize for the large mapping space needs to be payed when calculating the quartiles, which is outside the usual workflow of classification of a problem and finding its engineering difficulty.
In some embodiments, the method further comprises the measure of evaluation of the mapping results based on software metrics, wherein the software metrics measure the completeness selectivity of the mapping. The software metrics do not measure the underlying artifacts. This measure provides a check whether the mapping is useful to the application.
In some embodiments, the distribution calculation of the classification space is based on a calculation of weighted multi-selection in technical boundary taxa and a calculation of distribution and quartiles. This allows users to express priorities by weighing of choices.
In some embodiments, a weight is placed on the selection of a technical boundary condition based on technical requirements, the weight is multiplied with the weight of the combination created by the multi-selection and used the resulting value to look up the quartile in the distribution. Quartiles are not selected explicitly but provide a mapping trick to reduce complexity. The distribution calculation is influenced by the dimension of the weighted selections. For example, all 9 dimensions are weighted, so it takes with 4 weights, meaning 4⁹times longer to calculate the distribution with standard algorithms for the single selection case. With multi-selections, the number of refinements goes to the exponent. We get to combination numbers of 10²⁸or 10³³, compare the above-described embodiments.
In some embodiments, a weight is placed on the relation between the engineering disciplines and concerns and the products or services development requirements, and the relation weight is used as a count absorbed by the weight calculation. The calculation of the actual engineering complexity stays as small and fast as it was with pure multi-selection in both cases.
In some embodiments, the method allows a combination of both priories the mapping and the selection is allowed, depending on users' preference. For example, the placing of the weights influences the rating of e.g. engineering difficulty.
In some embodiments, the distribution calculation results are stored and evaluated for further subjecting the calculation results to a metric based ranking.
In some embodiments, the method further comprises the measures of processing an executable, which performs a distribution calculation of the classification space, wherein the distribution calculation of the classification space is at least based on a distribution and quartiles, wherein the process of the executable comprises the steps of calculation of the combination vectors at system start, reading mapping data and calculation probabilistic distribution and quartiles as well as publishing new distribution to engineering goal calculation. Segments meaning products or services or their development requirements can be described by several segments. The engineering goals are calculated by an integrated executable which calculates the underlying distribution. The calculation may be implemented in Excel in an adapted VBA script providing an achieved speed up to about 3 seconds compared to several minutes. No more system interruptions have to be dealt with and much higher stability is provided.
In some embodiments, the reading step and the publishing step are separated. The separation of the steps allows the user to check the quality of the new distribution before using it. In the example of the above mentioned VBA script there is a hidden performance improvement, since all excel calculation and updating can be switched off during the reading step. This could take minutes depending on size and complexity of the excel workbook during which excel is unresponsive and instable.
In some embodiments, the distribution calculation algorithm calculates the probability of a value in the distribution of a number of combinations and the quartiles underlying the engineering difficulty calculation are derived from the accumulated probabilities for each value. Instead of the occurrence of a value for each possible selection, which could be up to 2.2 millions, the algorithms may be written in VBA and C++. For example, 512 combinations, representing the permutation of a selection vector (0/1) size n instead of approximately 2.2 million has the same result in distribution and quartiles, while n is the dimension of the solution space.
In some embodiments, the distribution calculation of the classification space is based on a calculation of a multi-selection in technical boundary taxa and a calculation of distribution and quartiles.
In some embodiments, the method further comprises the measures: normalization of the selection-counts, especially the multi-selection counts, creation of value-probability-pairs to calculate the probability of a value in the distribution calculation and a distribution calculation from these value-probability-pairs. This provides the advantage of reduction of size of the frequency distribution. It's the n-time the normalization value instead of the product of maximum relation counts for each segment, while n is the number of dimensions of the solution space, e.g. n=9. The creation of value-probability-pairs especially is done for each segment. These pairs can be calculated in a linear algorithm from the normalized choice combinations in less than 5 ms in C++. The values add up the probabilities multiply over the segments. The resulting probability is added to the respective value in a frequency distribution. As the value-probability-pairs are a lot less that the original relation counts or normalized relation counts for each segment, the algorithm is orders of magnitude faster. In C++ for example, single-threaded the calculation time is unnoticeable, less than 20 ms. In excel VBA e.g. it is below 1 second.
In some embodiments, the method further comprises the measures: count data for frequency distribution per segment, prioritization of combinations to segment frequency distributions, combination of segment frequency distributions to a classification space frequency distributions. Frequency distribution may be displayed in histograms.
In some embodiments, the computer-implemented method further comprises the measure of horizontal stripping, wherein blocks of the solution space are given to separate threads by provided disjunct index subsets, therefrom resulting buffered frequency distributions are merged into the combined result in parallel by providing disjunct frequency distributions' values to the threads. For example, a speedup is reached by the horizontal striping. Blocks of an n-dimensional solution space, in a preferred embodiment n=9, are given to separate threads by provided disjunct index subsets. The resulting buffered histograms are merged into the combined result in parallel e.g. by providing disjunct histogram values to the threads such that no synchronization is needed.
In a CPU based example, with current hardware 4 Kernels, the FPUs Speed up against single threaded is up to 25%. In a GPU based example, based on applied massive parallelism (amp), a 32 bit library doesn't scale as necc. For precise calculation 128 bit library is necessary. For validation, a multi-precision library is used with the advantage that number sizes can be set in the program as needed to overcome the fixed number sizes by usual programming languages and math libraries. Block based, like CPU, but with 50 to 100 time more threads like CPU-memory overhead makes it slow. Laptop GPUs, for example share their main memory with CPU and therefor are slower.
In some embodiments, the method further comprises the measure of vertical stripping, wherein two or more segments' frequency distributions are combined to a combined frequency distribution, which combined frequency distribution further is reduced to a value-probability-pair frequency distribution and wherein these combination and reduction steps are repeated until all product or service development segments are frequency distributions are combined. As the combined histograms are very sparse the reduce step eliminates one or two orders of magnitude in input to the next combination step the overall algorithm is several orders of magnitudes faster than horizontal striping. For example, with 5 prioritization weights and 1000 as normalization factor it is less than 50 ms to almost estimated 7000 days or 206200 evaluated combinations to 1.32915.10¹⁴evaluated combinations, see FIG. 6. For example, the vertical striping algorithm can be implemented directly in VBA as is uses neither multi-threading nor GPU calls.
In some embodiments, the distribution calculation results are stored and evaluated for further subjecting the calculation results to a metric based ranking.
In some embodiments, there is a computer system for automatic derivation of software engineering artifacts, comprising a first subsystem with the components: a classifier software component for the classification of the technical boundary conditions, a calculation software component for the deduction of the technical requirements, a first mapping software component for mapping the technical requirements to engineering artifacts, an I/O-component for receiving the technical boundary conditions data and for providing the calculation results, a storage component and a second subsystem which provides a distribution calculator software component for the distribution of the classifications, wherein the distribution calculator software component for the distribution of the classifications provides a calculation of multi-selection and a calculation of distribution and quartiles.
In some embodiments, the distribution calculator software component for the distribution of the classifications provides a calculation of weighted multi-selection combinations and a calculation of distribution and quartiles.
In some embodiments, the storage component comprises at least a data base containing relevant data for the mapping processes.
In some embodiments, the distribution calculator software component has access to the storage component for storing the calculation results.
In some embodiments, the first subsystem comprises an evaluation software component for subjecting the calculation results to a metric based ranking.
In some embodiments, the computer system provides an executable performing a distribution calculation of the classification space, providing the distribution calculation of the classification space, which is at least based on a distribution and quartiles, wherein the executable comprises the further software components for calculation of the combination vectors at system start, reading mapping data and calculation probabilistic distribution and quartiles, publishing new distribution to engineering goal calculation.
In some embodiments, the computer system is comprised in one system device instead of several subsystems. This results in an advantage of speed up to about 3 seconds compared to several minutes. No more system interruption are to be dealt with and higher stability is provided. The need for an external tool is eliminated. With this approach, we no longer rely on advanced hardware, that would be limited to a couple of orders of magnitude. The presented method and system allow a speed up of 23 orders of magnitudes.
As another example, some embodiments include a computer program product is claimed, having program instructions which, when the program is executed by a computer, cause the computer to carry out the steps of the methods described herein. Furthermore, a provision apparatus for storing and/or providing the computer program product is a data storage medium that stores and/or provides the computer program product, for example. In some embodiments, the provision apparatus is a network service, a computer system, a server system, particularly a distributed computer system, a cloud-based computer system and/or a virtual computer system that stores and/or provides the computer program product preferably in the form of a data stream, for example.
This provision is effected as a download in the form of a program data block and/or instruction data block, e.g. a file, particularly as a download file, or a data stream, particularly as a download data stream, of the complete computer program product, for example. This provision can alternatively be affected as a partial download that consists of multiple parts and is downloaded particularly via a peer-to-peer network or provided as a data stream, for example. Such a computer program product is read in, for example using the provision apparatus in the form of the data storage medium, in a system and executes the program instructions, so that the methods described herein may be executed on a computer, or configures the creation device such that it creates cited system and/or execution unit.

Definitions

As used herein, the term technical system refers, for example, to a device, apparatus, or a plant. A technical system can, for example, be a field device, a generator or a power plant, e.g. a wind turbine, a solar power plant or a water-power plant. In some embodiments, the technical system comprises a plurality of hardware components and/or software components. Furthermore, the technical system can, for example, comprise at least one component having a communication interface configured to connect the apparatus and/or a test environment.
Unless specifically stated otherwise, as apparent from the following discussions, it is appreciated that throughout the specification discussions utilizing terms such as “processing”, “computing”, “computer-based”, “calculating”, “determining”, “generating”, “configuring” or the like, refer to the action and/or processes of a computer that manipulates and/or transforms data into other data, said data represented as physical, e.g. such as electronic, quantities. The term “computer” should be expansively construed to cover any kind of electronic device with data processing capabilities, including, by way of non-limiting example, personal computers, servers, handheld computer systems, Pocket PC devices, Cellular communication device and other communication devices with computing capabilities, processors and other electronic computing devices.
As used herein, the term “processor” refers, for example, to controllers, microcontrollers (e.g. digital signal processor (DSP) possibly in combination with memory and storage units memory unit storing computer-readable instructions, application specific integrated circuit “ASIC”, etc.), processor cores, central processing units, integrated circuits/processing circuitry (e.g. application specific integrated circuit “ASIC”, field programmable gate arrays “FPGA” etc.) or digital signal processors. Furthermore, the term “processor” can, for example, refer to virtual processors, virtual CPUs, soft processors or soft CPUs. Moreover, said processor can, for example, be configured to execute computer readable instructions such that said processor be configured to perform functions which implement the teachings herein.
As used herein, the term “module” refers, for example, to a processor and/or a memory unit storing computer-readable instructions. For example, the processor is specifically configured to execute the computer readable instructions such that said processor is configured to perform functions which implement the methods herein, such as a step of the method. Furthermore, the term “module” can, for example, refer to means (e.g., a processor) which are configured to implement/execute functions/steps of the methods.
As used herein, the term “subtree”, “tree” or the like, refer, for example, to data structures storing information about the technical system and/or components of the technical system. In some embodiments, a subtree is a branch of a (component fault) tree or a partial (component fault) tree, defined, for example, by a selected node in the tree as top level node of the subtree.
As used herein, the term “model”, “component”, “failure mode” and other elements of a component fault tree or the like, refer, for example, especially in conjunction with a subtree or tree to data structures containing information about the technical system and/or its components.
As used herein, the term “acquisition module” refers to a sensor or measurement equipment to measure a physical quantity. For example, an acquisition module can be a LIDAR to measure upcoming guests of wind and/or an acceleration sensor to measure the acceleration of the wind turbine and/or a speed sensor to measure a rotor speed of the wind turbine and/or a pitch angle sensor to measure a pitch angle of blades of a wind turbine and/or a power sensor to measure generated electrical power of a wind turbine and/or a speed sensor to measure an actual wind speed driving the wind turbine.
In some embodiments, the systems and/or methods are implemented by a processor and/or a memory device unless otherwise noted. In detail, to implement and/or execute the methods, components, devices etc. comprise at least one processor and/or at least one memory device unless otherwise noted. Additionally, the method, components, devices etc. comprise, for example, other features known by a skilled person. For example, these features can be an input device, like a computer mouse, or a display device, like a TFT-display.
In some embodiments, the method for successful implementations of projects, product, or service developments or even of new business models demands contains a scalable and reproducible deduction of technical concerns. The provided semi-automated method for example helps to define cornerstones of an engineering strategy by determining the minimal needed quality of core artifacts of the engineering and operation process, like requirements, enterprise architecture, source code, test strategy, test plan, change requests, etc. These artifacts can be part of an engineering canvas, for example comprising several building blocks like requirements management, architecture management, etc., see Tab. 1.
In some embodiments, the method comprises two main processes. On the one hand, the process definition. First, there are basic definitions like roles, artifacts, engineering concerns, building block refinements, representing the taxa for classification, to be considered. Further, there is an initial concept or canvas concerning a product or service idea, there are relationships, mapping to engineering concerns, software engineering canvas, and an artifact role mapping. These are complemented by a validation of the definition and mappings, e.g. completeness and variance.
Finally there could be an evaluation by examples. On the other hand, the method comprises the process Goal derivation. In tabular 1 for example, a results overview of the top-level engineering goals is shown.

TABLE 1

Example of Interface: Assignment Software Engineering
Building Blocks (SEBB) to quality levels

Requirements	Architecture	Software		Delivery &	Operations
management	management	implementation	Test	Deployment	management	Maintenance

1.5	2	2	2	2	2	2		Basic
								Engineering
								Quality
4	4	4	4	4	4	4		Peak
								Engineering
								Quality
2	4	2	1	2.5	4	4	Min	Automation
								grade
3	4	4	4	4	4	4	Max	Automation
								grade
2	4	3	3	4	4	4	Min	3
4	4	4	4	4	4	4	Max	Estimation
								capability

Needed quality of an artifact means that it has to contain specific information items and, depending on the quality levels, quantitative data that helps to control the development process and the maturity and completeness of the artifacts. In some embodiments, the method may be focused on artifacts, not on the process how these artifacts are created and maintained and can therefore be used regardless of the underlying development process.
In a first step, technical boundary conditions are identified and classified. Then, the technical requirements are deduced in an automated software-based process. The relations between specific building block refinements of the initial concept con and the requirement types req are defined and justified by a so-called Mapping map-con, see FIG. 3. This step is based on a model that allows selecting the major characteristics of new projects, product or service developments or even of new business models. This model can be implemented in Microsoft Excel. For users' selections, single selections in drop down boxes are preferably provided.
The deduced technical requirements can be classified in requirement types that have an impact on implementation, operation of products and services or even the design. Concerning the before mentioned quality levels, an overall requirements severity for a project, product or service development is calculated based on this selection. This requirements severity level for example is between 0 and 4 and defines the minimum quality to be achieved, and therefore the minimum content that the key artifacts of the project, product or service development have to contain. Additionally, suggestions for quality levels can be calculated for the distinct key artifacts of an engineering project.
The requirements, a test plan, source code etc can define different impacts on the needed content and quality of the distinct artifacts. Specifying the required quality levels of the key artifacts that are for example organized along major phases of a software development project, can be based on a software engineering template.
In a further step of this automated software-based process, these technical requirements are mapped to engineering artifacts and concerns, see FIG. 3, map-con. For the automated calculation of quality levels, a number of mappings are defined in the before mentioned model. Different Requirement Types req are derived from different content and quality that is needed for the artifacts of a project. E.g. the requirements complexity is much higher when dealing with a complex multi-level project. The mapping of the requirement types for example to a software engineering template defines a minimum quality level for the key software artifacts based on the characteristics of the requirement types. In an advantageous embodiment of the invention the deduced engineering artifacts are further mapped to responsibilities.
This process can be followed by an evaluation step, see Tab.2, that is based on software metrics. In an embodiment, the evaluation results are used to adapt the initial classification of the technical boundary conditions in several iterations.

TABLE 2

Definition process

Basic			Plausibility
definitions	Mappings	Distribution	checks
				Evaluation

Artifacts	Concept-	Concept-	Complete-	Test
Roles	to-	and	ness	scenarios
Concept	Engineering	Engineering	Statistics
elements	concerns	goal
Engineering	Software	calculation
concerns	engineering	on EC
Metrics	template	(in second
(in first	Artifact-	subsystem 2)
subsystem 1)	quality-
	mapping
	Artifact-
	automation-
	mapping
	Valuation
	mapping
	Collaboration
	mapping
	(in first
	subsystem 1)

As shown in FIG. 3, the model does not draw a direct relation between an initial concept and the software engineering template SEC, but uses so called requirements types req to decouple the two sides in order to reduce the effort for defining and describing the relationships between. Furthermore, the requirements types req help in narrowing the semantic gap. A major benefit is the refinement of the problem space and an improvement of the mapping characteristics. The requirement types req are kind of a middle-tier that facilitates relating a development concept with a software engineering canvas SEC. The requirement types req classify a software or system service or a software product from various key demands, like functional suitability, functional quality, engineering quality, and operational quality. Depending on the importance of the individual requirement types req this has an impact on the way development and operations of a service have to be carried out.
The defined mappings, map-con, map-reg, relate every element of the method to each other, see FIG. 3. In the first place, there is a concept, con, for example a project idea, a development plan, a business model canvas, which comprises several Building blocks BB or Building Block Refinements BBR. In the concept-to-requirements-mapping, map-con, technical requirements req are deduced in an automated software-based process.
The concrete form of the concept con, how complex or multi-sided it is, has an impact on the requirement types, e.g. requirements' complexity, resulting in different content and quality that is needed for the artifacts of a project, compare Tab.3. Both mappings, map-con, map-reg, for example provide an automated calculation of the quality levels respectively.
In the requirements-to-engineering-artifacts-mapping, map-req, technical requirements are mapped to engineering artifacts. In the field of software engineering, this mapping defines a minimum quality level for the key software artifacts based on the characteristics of the requirement types. The requirements-to-engineering-artifacts-mapping, map-req, for example results in software engineering building blocks SEBB. With respect to an underlying software engineering canvas SEC and the derived software engineering building blocks SEBB core artifacts CA are obtained. For each key software or role artifact CA and for each quality level respectively, the necessary roles for developing and maintaining the artifact as well as which roles should use the artifact in order to be able to fulfil the tasks of the role are defined.
Generally, any mapping is explicitly specified and can be modified, e.g. in the model, in order to better reflect the needs of a specific organization or division. Such modifications would also allow, to remove or add requirement types, to change the impact of the concept elements on the requirement types, to change the minimum required quality level for key software artifacts, to add or remove key artifacts, and to redefine the required roles for each artifact.

TABLE 3

Automated derivation process

Concept	Engineering	Artifact related
Classification	difficulty	goals

Select on entry of	Calculate the	Quality
each segment of	value for each	Automation grade
the initial concept	engineering	Evaluation
of project idea,	concern map-con,	capability
development	and relate using	Artifact
plan, business	the distribution	collaboration
model canvas	of all possible
	elections.
	Aggregate the
	ratings over all
	engineering
	concerns
	(Average)

In some embodiments, the method is suitably executed by a computer system, see FIGS. 1 and 2, comprising a classifier software component, class, for the classification of the technical boundary conditions, a calculation software component, map-con, for the deduction of the technical requirements, req, and at least one mapping software component, map-req, for mapping the technical requirements to engineering artifacts, disciplines and concerns. The computer system further comprises an I/O-component, I/O, for receiving the technical boundary conditions data and for providing the calculation results, especially for the input of classification and mapping data and for the output of engineering goals and respective mappings.
In some embodiments, the computer system comprises at least a storage component, stor, the storage component e.g. comprising at least a data base containing relevant data for the mapping processes. In the storage component, stor, basic definitions, mappings and distributions are stored. In an advantageous embodiment of the invention, these components are combined in a first subsystem. In some embodiments, a second subsystem provides a distribution calculator software component for the distribution of the classifications, which is a calculation of single-selection combinations and/or a calculation of distribution and quartiles. The distribution calculation results are stored in the storage component. In some embodiments, the first subsystem comprises an evaluation software component for subjecting these results to a metric based ranking. The I/O-component provides scalable and reproducible calculation results.
FIG. 4 shows a diagram for the distribution calculation DC interaction with the interface module I/O and storage module, stor, of the first subsystem 1. In some embodiments, the interface component I/O provides input of classification and mapping data to the distribution calculation DC in the second subsystem 2. The distribution calculation module DC uses the mapping, map-con, between concept and engineering concerns to calculate a distribution of all possible classifications, req. This resulting distribution is created and stored to the datastore, store. In the calculation of the distribution and quartiles, all combinations can be enumerated using an odometer with a wheel for each segment of the concept, con. The number of digits per wheel are the number of selections in a segment. The value of the digit holds or refers to the value of a selected refinement. The value of a combination is the sum of the values of all selected refinements. For all combination values, a histogram of the distribution is created for the quartiles to be calculated easily. The overall number of combinations is the product of all segment combinations. We create a distribution for each engineering concern. The interface component I/O further provides the output of engineering goals and mappings to a user.
In some embodiments, the system provides an automated mapping between a concept, con, and artifact-based engineering goals. Selections in the concept segments are provided to the system as input data. Output, for example, are artifact-names with quality grade, automation grade and estimations capability. Grades are aggregated at discipline and organization level to provide an overview.
The main challenge of the described method is to bridge the semantic gap between concept and engineering artifacts in a comprehensible way for the definition of the mapping and the evaluation of a concept selection. Therefor mappings between intermediate values are provided, like the concept-to-requirements-mapping, map-con, and the requirements-to-engineering-artifacts-mapping, map-req, but also further mappings from a software engineering canvas, SEC, to artifact quality, automation grade and/or evaluation capability. The methods described herein include a fully automated calculation, which limits are based on calculated quartiles and not randomly defined. To conclude, the presented computer-implemented method for successful implementations of projects, product or service developments or even of new business models demands, contains a scalable and reproducible deduction of technical concerns. After the identification and classification of the technical boundary conditions, in a first step, technical requirements are deduced in an automated software-based process. In a further step of this automated software-based process, these technical requirements are mapped to engineering artifacts and concerns. In some embodiments, the deduced engineering artifacts are further mapped to responsibilities. This process can be followed by an evaluation step that is based on software metrics. In some embodiments, the evaluation results are used to adapt the initial classification of the technical boundary conditions in several iterations.
In some embodiments, the method is suitably executed by a computer system comprising a classifier software component for the classification of the technical boundary conditions, a calculation software component for the deduction of the technical requirements and at least one mapping software component for mapping the technical requirements to engineering artifacts, disciplines and concerns. In some embodiments, the computer system further comprises an I/O-component for receiving the technical boundary conditions data and for providing the calculation results. The computer system comprises at least a storage component, the storage component e.g. comprising at least a data base containing relevant data for the mapping processes. In some embodiments, the components are combined in a first subsystem. In some embodiments, a second subsystem provides a distribution calculator software component for the distribution of the classifications, which is a calculation of single-selection combinations and/or a calculation of distribution and quartiles. The distribution calculation results are stored in the storage component. In some embodiments, the first subsystem comprises an evaluation software component for subjecting these results to a metric based ranking. The I/O-component provides scalable and reproducible calculation results.

LIST OF REFERENCE SIGNS

EC Engineering goal calculator/calculation
DC Distribution calculator/calculation, Definition
stor Data storage
dis Distributor, e.g. project distribution
I/O Input/Output, user interface
clas Classification, e.g. project classification
pre presentation
agg aggregator
req requirements, e.g. engineering requirements
art artifacts, e.g. quality, automation, grade, valuation, collaboration
1 first subsystem
2 second subsystem
con concept, e.g. project idea, development plan, business model canvas, comprising several Building blocks (BB)
BBR Building Block Refinements
SEBB Software Engineering Building Blocks
map-con concept-to-requirements-mapping: technical requirements are deduced in an automated software-based process
map-req requirements-to-engineering-artifacts-mapping: technical requirements are mapped to engineering artifacts
CA core artifacts
SEC software engineering template or canvas

Claims

1. A computer-implemented method for automatic derivation of attributes of software engineering artifacts, which attributes arise from technical boundary condition of products or services, the method comprising:

deducing technical requirements based on classifications of the technical boundary conditions;

mapping the deduced technical requirements of the artifacts to engineering disciplines and concerns;

mapping the calculated engineering artifacts to responsibilities;

adapting the classification of the technical boundary conditions based on the evaluation results in iterations; calculating a distribution of the classification space based on a calculation of a multi-selection in technical boundary taxa; and

calculate distribution and quartiles.

2. A computer-implemented method according to claim 1, wherein mapping of the deduced technical requirements of the artifacts to engineering disciplines and concerns at first is kept between a segment choice of the technical requirements and at second for each multi-selection possibility per segment, the relations to all engineering disciplines and concerns are counted.

3. A computer-implemented method according to claim 1, further comprising:

evaluating a specific multi-selection, wherein the relationships, built in the mapping step, are counted and related to the distribution of possible multi-selections split up into quartiles; and

selecting a quartile based on technical requirements, wherein the quartile definition is based on a calculated distribution which entails an algorithm considering all selection possibilities of the classification space.

4. A computer-implemented method according to claim 1, further comprising evaluating the mapping results based on software metrics measuring completeness selectivity of the mapping.

5. A computer-implemented method according to claim 1, wherein the distribution calculation of the classification space is based on a calculation of weighted multi-selection in technical boundary taxa and a calculation of distribution and quartiles.

6. A computer-implemented method according to claim 1, further comprising placing a weight on the selection of a technical boundary condition based on technical requirements;

wherein the weight is multiplied with the weight of the combination created by the multi-selection and used the resulting value to look up the quartile in the distribution.

7. A computer-implemented method according to claim 1, further comprising placing a weight is placed on the relation between the engineering disciplines and concerns and the products or services development requirements;

wherein the relation weight is used as a count absorbed by the weight calculation.

8. A computer-implemented method according to claim 1, further comprising storing the distribution calculation results and evaluating the results for further subjecting the calculation results to a metric based ranking.

9. A computer system for automatic derivation of software engineering artifacts, the system comprising:

a first subsystem with the components;

a classifier software component for classification of technical boundary conditions;

a calculation software component for deduction of technical requirements;

a first mapping software component for mapping the technical requirements to engineering artifacts;

an I/O-component for receiving technical boundary conditions data and for providing the calculation results;

a storage component; and

a second subsystem providing a distribution calculator software component for the distribution of the classifications;

wherein the distribution calculator software component for the distribution of the classifications provides a calculation of multi-selection and a calculation of distribution and quartiles.

10. A computer system according to claim 9, wherein the distribution calculator software component for the distribution of the classifications provides a calculation of weighted multi-selection combinations and a calculation of distribution and quartiles.

11. A computer system according to claim 9, wherein the storage component comprises at least a data base containing relevant data for the mapping processes.

12. A computer system according to claim 9, wherein the distribution calculator software component has access to the storage component for storing the calculation results.

13. A computer system according to claim 9, wherein the first subsystem comprises an evaluation software component for subjecting the calculation results to a metric based ranking.

14. A non-transitory executable computer program product comprising instructions which, when the program is executed by a computer, cause the computer to:

deduce technical requirements based on classifications of the technical boundary conditions;

map the deduced technical requirements of the artifacts to engineering disciplines and concerns;

map the calculated engineering artifacts to responsibilities;

adapt the classification of the technical boundary conditions based on the evaluation results in iterations;

calculating a distribution of the classification space based on a calculation of a multi-selection in technical boundary taxa; and

calculate distribution and quartiles.