LUND UNIVERSITY

PO Box 117221 00 Lund+46 46-222 00 00

The 4+1 View Model of Industri–Academia Collaboration

Runeson, Per; Minör, Sten

Published in:[Host publication title missing]

DOI:10.1145/2647648.2647651

2014

Link to publication

Citation for published version (APA):Runeson, P., & Minör, S. (2014). The 4+1 View Model of Industri–Academia Collaboration. In [Host publicationtitle missing] Association for Computing Machinery (ACM). https://doi.org/10.1145/2647648.2647651

Total number of authors:2

General rightsUnless other specific re-use rights are stated the following general rights apply:Copyright and moral rights for the publications made accessible in the public portal are retained by the authorsand/or other copyright owners and it is a condition of accessing publications that users recognise and abide by thelegal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private studyor research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal

Read more about Creative commons licenses: https://creativecommons.org/licenses/Take down policyIf you believe that this document breaches copyright please contact us providing details, and we will removeaccess to the work immediately and investigate your claim.

The 4+1 View Model of Industri–Academia CollaborationPublished in WISE’14, http://dx.doi.org/10.1145/2647648.2647651

Per Runeson and Sten Minor

Abstract

Industry–academia projects exist in complexcontexts of various stakeholders, time perspec-tives, and goals. In order to analyze projectsand communicate about them, we have de-fined an “architectural” model for industry–academia collaboration, inspired by Kruchten’ssoftware architecture model. The model hasfour views of i) time, ii) space, iii) activityand iv) domain, corresponding to the ques-tions: when, where, how and what. The +1view is the scenario, binding the other four to-gether. We illustrate the model by applyingit to the Industrial Excellence Center EASEand the Sigrun Software Innovation and En-gineering Institute. The model helps analyz-ing industry–academia collaboration projects,to find gaps and reduce redundant work.

1 Introduction

When defining industry–academia collabora-tion projects, there is often an urgent needto communicate not only what topics to beworked on, but also how, in which time frame,and by whom. Further, whether a projectshould be funded or not is not only a mat-ter of topic, but also how it may be comple-mentary to existing projects and how it en-ables collaboration between new stakeholders.In order to support this communication and as-sessment of ongoing collaboration projects, wehave defined the 4+1 view model of industry–academia collaboration. The model is based

on experiences from more than a decade ofcollaboration between industry and academiain different projects, and is of course inspiredby Kruchten’s 4+1 View Model of Architec-ture [5]. More specifically, this work is basedon experiences from the Industrial ExcellenceCenter for Embedded Applications SoftwareEngineering (EASE)1 and our forming of theSigrun Software Innovation and EngineeringInstitute2.

Previously presented models for industry–academia collaboration focus on activities, asproposed by Gorschek et al.[4], or relations,as proposed by Sandberg et al. [8]. Otherwork includes experience reports on specificprojects [7] and summaries of challenges forthe industry–academia collaboration [9]. Themodel proposed in this paper addresses the“architecture” of industry–academia collabora-tion, i.e. the “components” and their struc-tural relationships.

We present the model in section 2, and intro-duce each of the 4+1 views in subsections 2.1to 2.5, elaborate an example from EASE andSigrun in section 3, while section 4 concludesthe paper.

2 4+1 View Model Overview

The proposed the 4+1 view model of industry–academia collaboration consists of four views,

1http://ease.cs.lth.se2http://www.sigrun.se. The organizational host for

Sigrun is currently under investigation, but for simplic-ity, we refer to it as a separate unit here as it has beenup till now.

Page 1

according to the traditional when, where, whatand how questions. The +1 view, is like inKruchten’s model, the scenario instantiation,which binds them all together.

1. Time view (when)

2. Space view (where)

3. Activity view (how)

4. Domain view (what)

5. + Scenario view

The model is graphically represented in Fig-ure 2, and is elaborated below.

2.1 Time View – When?

In the forming of any project, the time con-straints are by definition important, since aproject can be defined as “a temporary en-deavor undertaken to create a unique product,service, or result” [1]. This is of course appli-cable for an industry–academia project as well.However, here we refer to target time for thejoint project. Is the project aiming at address-ing current problems, next product generationproblems, or a future problem, which we onlysee at the horizon from an industrial point ofview? We therefore define the time view interms increasing time to when the research isexpected to be practiced:

• Now – best current practice

• Soon – next practice

• 3–5 years – applied research

• 5+ years – basic or fundamental research

Now, there are best software engineeringpractices in industry, which most probablycan be shared between industry partners.Academia may have a catalyst role is this peer-to-peer exchange, but more often, this is therole taken by consultancy companies. Next

is about practices at the door-step, for exam-ple practices developed and evaluated in pi-lot projects. The industry–academia collabo-ration may here involve scaling up the appli-cation of the practices, but gradually, indus-try or industry consultants take the responsi-bility for the roll-out. The 3–5 years perspec-tive is fruitful for industry–academia collabo-ration, where the slower, more thoughtful paceof academia processes may complement the in-dustry higher speed pace. Industry relevant,long-term challenges may be addressed, stillwith continuous interaction between industryand academia, and incremental delivery of re-sults. Finally, basic research on software engi-neering should also be driven by industry rel-evance, however, not for the current industrialstate, but for a future situation. Continuousinteraction with industry is important also inbasic research, but the time to perspective toexpected application is much longer – and thusmore risky.

2.2 Space View – Where?

Any type of industry–academia collaborationrequires continuous interaction. Software en-gineering researchers need influence from in-dustry practice to ensure relevance, and in-dustry practitioners need to adopt new in-fluxes from research to improve the practice.Even though electronic communication meansalso support the industry–academia communi-cation, the face-to-face meeting is crucial inthis kind of collaboration, since the communi-cation is about getting to know each others per-spectives, trying out novel ideas, and also buildup trust. Therefore, the spatial view is impor-tant, i.e. where the researchers and the indus-try practitioners are located plays a significantrole in how tight the industry–academia col-laboration may become. This issue very muchboils down to how much extra time either partyhas to spend on traveling for a meeting, whichbecomes an overhead cost. We therefore distin-guish between four grades of spatial vicinity:

• Local – almost no traveling time

Page 2

when% where%

what% how%

Figure 1: Overview of the 4+1 view model.

• Regional – traveling time of 1-2 hours, i.e.a meeting takes at least half a day

• National – traveling of 2+ hours, i.e. anymeeting takes a full day

• International – traveling takes more thanone day

How far these zones reach depend on wherethey are located. For a research center in Lux-embourg, for example, the whole country canbe defined as local, while in a US context, thenational level would rather be referred to as thestate level. In our specific context, as shown inFigure 2(top-right), local is the cities of Lundand Malmo, regional is the southern Sweden,national is Sweden, and international is Europeand beyond.

2.3 Activity View – How?

The activities undertaken in industry–academia collaboration may take manydifferent forms. All collaboration projectsmust not have the goal of finding the “softwareengineering silver bullet” – in fact not oneshould have – but project may have threeprincipal goals:

• Networking – create contacts between ac-tors

• Catalyzing – initiate activities

• Executing – running joint research and im-provement activities with specific goals

We have identified four principal types of ac-tors, which are involved in these kinds of col-laboration projects, below presented with ex-amples.

• Financing – science or innovation fundingagency, companies, or independent funds

• Knowledge provider – university, researchinstitute, or company

• Service provider – consultancy company,which transfers knowledge and experience

• Product provider – software develop-ment company, that ultimately wants theknowledge

A networking activity may be an indus-try seminar, where product providers andknowledge providers meet and start talking toeach other. From these, knowledge exchange

Page 3

projects or internal improvement projects maybe catalyzed. What we normally refer to as“research projects” fall in the executing cat-egory. The need for financing is very smallfor networking and catalyzing activities, whilethe execution of larger projects require externalfunding, or co-funding.

2.4 Domain View – What?

Finally, we zoom in on what to do within anindustry–academia project. First, we find itfruitful to distinguish between three categoriesof topic areas related to software engineering:

• Technology – the software technology initself; languages, tools, frameworks

• Engineering – the systematic approach todevelop and evolve software systems

• Management – the management of soft-ware engineering projects, products andservices; the business aspects

Research on, for example, software regres-sion testing, may address technical issues ofhow to identify regression test cases from codechanges (technology)[3], may be about whichvisualization approach is feasible for test man-agers (engineering)[2], or may be about theeconomics of various regression test strategies(management).

Another dimension of the What? questionis the industry domain, in which the researchtakes place. Software engineering has theunique characteristics of being cross-domain,to a large extent. Many software engineeringchallenges are the same or similar across indus-try domains, as for example a survey of foursafety-critical domains show (robotics, trans-portation, automation, and aerospace) [6]. Thechallenges may come in different flavors, and atdifferent points in time, but there is definitelya potential to learn across industry domains.Therefore, we characterize the research with re-spect to the industry branch(es) in which it isconducted.

when   where  

what   how  

Example:  EASE  

Figure 2: Scenario of the 4+1 view model forthe EASE Industrial Excellence Center.

2.5 Scenario View

The +1 view basically connects the other four,into one instance for each industry-academiacollaboration project. The dimensions of themodel are not completely orthogonal (it may,for example, be hard to think of a basic re-search project, based on networking only) butthey are sufficiently independent to provide keycharacteristics to be used when defining andcommunicating industry–academia collabora-tion projects. Next section presents the EASEIndustrial Excellence Center and the SigrunSoftware Innovation and Engineering Instituteas examples.

3 Example

The EASE Industrial Excellence Center, aspresented in Figure 2, is a ten year programto execute applied research projects in the fieldof embedded applications software engineer-ing, which now has endured for five years.EASE is based on the long term collaborationbetween knowledge providers Lund University(LU) and Blekinge Institute of Technology(BTH) as well as four software-intensive com-panies with offices in southern Sweden: prod-uct providers: Sony Mobile, Ericsson, Axis,and service provider : Softhouse, all in the tele-com and mobile domain. The center operatesin the industry–academia ecosystem in south-

Page 4

when   where  

what   how  

Example:  Sigrun  

Figure 3: Scenario of the 4+1 view model forthe Sigrun Software Innovation and Engineer-ing Institute.

ern Sweden. It is mostly funded by compa-nies (1/2), academia (1/3) and Vinnova3 (1/6).Since some of the company funds come in termson industrial PhD students, and other in-kindsupport, parts of the execution takes place inthe product and service provider companies.

The research agenda that has been and iscontinuously developed in collaboration withthe industrial partners and has been focussedon two main areas and four themes: 1. En-gineering : i) analysis and assessment of ag-ile and open source software engineering prac-tices, and ii) models and tools to bridge gapsbetween information artifacts for requirementsand testing. 2. Technology : i) tools for ubiqui-tous interaction and configuration, and ii) im-plementation of speculative parallelization inweb browsers.

Based on the 4+1 view model of EASE, wecan see that the center only operates in onebranch of industry, and that there is a miss-ing link from applied research to next prac-tice. In order to bridge this gap, Sigrun wasfounded in 2010 to constitute a bridge betweenapplied software research and industrial prac-tice. Sigrun promotes openness by creatinga forum for open collaboration and exchangeof software, development and business experi-

3http://www.vinnova.se, The Swedish Governmen-tal Agency for Innovation Systems

ence, and software innovations. Research andinnovations are brought to practical use mainlythrough running innovation projects includingparticipants from academia, product compa-nies, consultant companies, and public orga-nizations. The goal of innovation projects is totry out new technology, processes, or services inpractice in an industrial setting. Sigrun is opento membership for large companies, small andmedium-sized enterprises, startups, and publicorganizations.

By starting Sigrun, more light-weightprojects may be launched, with short-termcommitments. That has enabled us to includenew actors, specifically small and medium-sized enterprises (SME), and to expand intoother industry domains. The 4+1 model inFigure 3 show that Sigrun is a complementaryactor to EASE.

On the other end of the time view, basic re-search is funded by other kinds of projects,e.g. the Synergies ICT Framework programon Software Engineering for Open Innovation4.This is a long-term basic research project,funded by the Swedish National Science Foun-dation. In the project proposal, the projectwas positioned using the time view of the4+1 model, see Figure 4, where Sigrun andEASE are mentioned alongside other projects,thereby helping the project funders to assesswhether the project is complementary to oroverlapping existing projects. Further, we hereadded the education dimension, to illustratehow the research project contributes to educa-tion, which has both long term and near prac-tice components.

4 Conclusions

Industry–academia collaboration projects maybe defined along many different perspective,with respect to what to study, which time hori-zon, and which actors are involved. In order tosupport communication and analysis of collab-oration projects and activities, we have defined

4http://serg.cs.lth.se/projects/synergies/

Page 5

Best practice

Next practice

Applied research

Funda-mental

research

Actor

industry

Sigrun

EASEENGROSSPIEp

ELLIIT

Time frame

now

soon

3-5 years

5+ yearsSYNERGIES

Educ

atio

n

Figure 4: Overview of Synergies project timeview, from the project application.

the 4+1 view model of industry–academia col-laboration. We demonstrate by example, thatthe model is useful in analyzing gaps in theindustry–academia ecosystem, and communi-cating goals when proposing new projects. Inour context, the model has been useful to ana-lyze and communicate industry–academia col-laboration, and we think the model may be use-ful for other industry–academia ecosystems aswell.

5 Acknowledgments

This work was funded by EASE, the IndustrialExcellence Center on Embedded ApplicationsSoftware Engineering.

References

[1] A guide to the project management bodyof knowledge (PMBOK guide) 3:rd edition.Technical report, Project Management In-stitute, 2004.

[2] E. Engstrom, M. Mantyla, P. Runeson, andM. Borg. Supporting regression test scop-ing with visual analytics. In Proceedings ofthe 2014 IEEE International Conference onSoftware Testing, Verification, and Valida-tion, pages 283–292, 2014.

[3] E. Engstrom, P. Runeson, andM. Skoglund. A systematic reviewon regression test selection techniques.

Information and Software Technology,52(1):14–30, 2010.

[4] T. Gorschek, P. Garre, S. Larsson, andC. Wohlin. A model for technology trans-fer in practice. IEEE Software, 23(6):88–95,2006.

[5] P. B. Kruchten. The 4+1 view model ofarchitecture. IEEE Software, pages 42–50,November 1995.

[6] J. Pedersen Notander, M. Host, andP. Runeson. Challenges in flexible safety-critical software development – an indus-trial qualitative survey. In J. Heidrich,M. Oivo, A. Jedlitschka, and M. T. Bal-dassarre, editors, Proceedings 14th Interna-tional Conference on Product-Focused Soft-ware Process Improvement (PROFES), vol-ume 7983 of Lecture Notes in ComputerScience, pages 283–297, Paphos, Cyprus,June 2013. Springer.

[7] P. Runeson. It takes two to tango – an ex-perience report on industry–academia col-laboration. In Testing: Academic and In-dustrial Conference - Practice and ResearchTechniques (TAIC-PART), pages 872–877,2012. (Best paper award).

[8] A. Sandberg, L. Pareto, and T. Arts. Ag-ile collaborative research: Action principlesfor industry-academia collaboration. Soft-ware, IEEE, 28(4):74–83, july-aug. 2011.

[9] C. Wohlin. Empirical software engineeringresearch with industry: Top 10 challenges.In First Intl. Workshop on Conducting Em-pirical Studies in Industry - An ICSE 2013Workshop, pages 43–46, 2013.

Page 6