Experience modeling with graphs encoded knowledge for construction industry

(1)

Any correspondence concerning this service should be sent to the repository administrator:

staff-oatao@inp-toulouse.fr

Identification number: DOI: 10.1016/j.compind.2015.02.004

Official URL:

http://dx.doi.org/10.1016/j.compind.2015.02.004

This is an author-deposited version published in:

http://oatao.univ-toulouse.fr/

Eprints ID: 13767

To cite this version:

Kamsu-Foguem, Bernard and Abanda, Fonbeyin Henry

Experience modeling

with graphs encoded knowledge for construction industry

. (2015) Computers in

Industry, vol. 70. pp. 79-88. ISSN 0166-3615

Open Archive Toulouse Archive Ouverte (OATAO)

OATAO is an open access repository that collects the work of Toulouse researchers and

makes it freely available over the web where possible.

(2)

Experience

modeling

with

graphs

encoded

knowledge

for

construction

industry

Bernard

Kamsu-Foguem

a,

*

,

Fonbeyin

Henry

Abanda

b

a_Laboratoire_Ge´nie_de_{Production/ENIT-INPT}_–_University_of_Toulouse,₄₇_Avenue_d’Azereix,₆₅₀₁₆_cedex,_Tarbes,_France

b_Oxford_Institute_for_Sustainable_Development,_Department_of_Real_Estate_and_{Construction,}_Faculty_of_Technology,_Design_and_Environment,_Oxford_Brookes University,Oxford,OX30BP,UK

1. Introduction

Inthepast,therehasbeennostructuredapproachtolearnfrom constructionprojectsoncetheyhavebeencompleted.Nowadays, theconstructionindustryisadoptingasustainablestrategybyusing knowledgemanagementconceptsandtechniquestoimprovethe organizationofknowledgefromcompletedprojects[9].Knowledge canbe seen as the means to extractand capture the available informationwhichcanbeusedtoprovidebetterandefficientways ofimprovingprocesses,producingand/ordeliveringconstruction products (e.g., buildings). For instance, experiences that relate lessons learned of how well a product could be produced, maintained, used and so forth [12]. Knowledge-based quality

improvementinvolvesexperiencefeedbackmethodologythatcan beappliedwiththreekeymethodologicalelements:(a)analysis withasetoftools;(b)astepwiseproblem-solvingmethodand(c) usingcommonmetrics[7].Withregardstothefirst,recommended toolsincludestatisticaltechniques(e.g.,designofexperimentsand regressionanalysis),andnon-statisticaltools(e.g.,flowchartsand fishbonediagrams,orboth).Astepwiseproblem-solvingoffersa systematicanalysisandimpactslearningandknowledgecreation and consists of the following steps: problem identification, diagnosis,solutiongenerationandimplementation.Furthermore, stepwiseproblem-solvingtechniqueprovidesmodelsforproblems visualizationandsystemicunderstandingoftheirmain character-isticsandthereasoningoptionsavailable.Theinteractionbetween continuous improvements, experience feedback processes and stepwiseproblem-solvingproceduresleadstoknowledgecreation from structured learning processes [34]. Common metrics aid coordinateknowledgecreationeffortsbyincorporatingandaligning Keywords: AECindustry Conceptualgraphs Domainontology Knowledgeformalization Visualreasoning ABSTRACT

TheArchitecture,EngineeringandConstruction(AEC)industryisbecomingincreasinglyknowledge intensive.Knowledgemanagementhasbeenhailedasanenablerfortappingthisknowledgetoimprove efficiencyintheAECindustry.Inthisstudy,themainconceptsandbenefitsofknowledgemanagement, relationshipsbetweenknowledgemanagementandcontinuousimprovement havebeenexamined. Furthermore, emphasis has been laid on knowledge management through experience feedback processeswhichconstitutevaluableassetsfortheAECindustry.Thisisdonethroughtheemploymentof ontologiesandgraph-basedreasoningoperationsinelicitingandvisualizingknowledgeconceptsinthe AECdomain.Theproposedapproachwhichconsistsoftwomainaspectsprovidestheopportunityto examinetheconceptualandontologicalknowledgemodelswithassociatedreasoning.Inthefirst,the conceptsofknowledgemanagement,theirsignificanceandapplicationintheAECfieldarediscussed. Theseconddealswithamethodologicalframeworkforthemodelingandreasoning inthedomain ontology.Tofacilitateautomation,anontologygraph-basededitor,ConceptualGraphsUserInterface (CoGui)wasusedtomodelknowledgeandencodedreasoningintheknowledgebase.CoGuiencodes knowledgeasconceptualgraphsandreasoningasgraphoperationsthatcanbevisualizedinalogically preciseway,basedondomainontologies.ItemergedthatCoGuicould beveryusefulinacquiring informationthatcanbeusedincollaborationwithotherstocontinuouslyimproveinformationsharing andre-use.AFrenchAECcompanylocatedintheSouthwestregionwasemployedasacasestudyin illustratingknowledgemodelingthroughtheexperiencefeedbackprocess.Foreachphaseofexperience feedback,actionsthatwereimplementedinthecompanyarediscussed.

* Correspondingauthor.Tel.:+33624302337;fax:+33562442708.

E-mailaddress:Bernard.Kamsu-Foguem@enit.fr(B.Kamsu-Foguem).

(3)

theproblem-solvingprocessaimatimprovinginteractionandoffer pragmaticsolutionsinworkrelatedsituations.Despitethemeritsof theafore-mentionedexperiencefeedbackmethodologies,common major challenges related to computing in civil and building engineering still exist.Tizaniand Mawdesley [32] reportedthat (a)enhancingdigitalinformationmodeling;(b)extendingprocess modelingtechnologies;(c)improvingdecision–implication analy-sis,(d)integratingbroaderaspectsofexperiencefeedbackprocesses arestillquitecommon.Furthermore,inthefieldofcivilengineering, studies about experience feedback tend to focus on structural aspectswithlittleinterestinpreliminarystagesoffacilitatingworks. In France, experience feedback is more common in the field of industrial engineering than in AEC. The main reason is the automation of production systems in industrial engineering, compoundedbythefactthatindustrialproductsoftenhavesimilar characteristics.

Ithasbeenhailedthatsolutionsformostknowledge manage-mentchallengeslieininformationandcommunication technolo-gies (ICT). The increasing availability and quality of ICT; for example,theinternet,underlying communicationandemerging technologies aretransformingthewaypeople shareknowledge andideas.Theeaseofuseandcharacteristicsofthesetechnologies have made them very popular in a very short period of time. However; compared to otherindustries, theAEC industry lags behindwithregardstothesuccessfulimplementationofICTandin tappingthepotentialofthenewand/orinnovativetechnologiesto improving productivity, achievinggreater efficiency and higher quality [22]. Thus, knowledge of products, services and/or processes frompast AECprojects arehardly fullyexploited for useinotherprojectsleadingtoinformationduplications, incon-sistenciesandinefficienciesindeliveringsuchprojects.Thisstudy aims to employ an ontology-driven approach to develop a knowledge management model for construction projects. The modelwillserveasaknowledge-baseddecisionsupportsystem. Tofacilitateunderstanding,thepaperisdividedintofoursections. Section 2 provides a background of the state-of-the-art about knowledgemanagementapplicationsintheconstructionindustry. Section 3 presentsan ontology-driven approach for knowledge modeling in the construction AEC industry. In Section 4, an illustrativecasestudyisemployedtodepictthemethodologyand concepts ofexperiencefeedback.Thediscussion andanalysisof majorissuescoveredinthismanuscriptarediscussedinSection

5.TheconclusionispresentedinSection6byawayofsummaryof thekeypartsofthepaper.

2. State-of-the-art

2.1. GenerationsofknowledgemanagementapplicationsintheAEC industry

Knowledgemanagementisrecognizedasakeyresourceanda crucialenablerforcontinuousprocessimprovementinconstruction

projects.Ithasbeen widelyacknowledgedthatoneof thekey sustainableadvantagesthatafirmpossessescomesfromwhatit collectivelyknows,howefficientlyituseswhatitknowsandhow readily it acquires and uses new knowledge. Organizational knowledge could reduce the time spent on problem-solving andincreasesthequalityofwork.Recent developmentsin ICT have prompted organizations to utilize platforms such as corporateintranetsandcollaborativeextranetsforcollaborative knowledgesharing.Companiesarealsoimplementingextranets toshareinformationofexperiencefeedbackscenariosexpressed forexplicitknowledgecapitalizationandexploitation.Inthese situations,experiencefeedbackisusefulintermsofhowwellthe organizationworksinmanyaspects,from learningknowledge fromexperiencefeedbackinformationandusingitinimproving management strategies. The preceding statement underscores thedifferencebetweeninformationandknowledgemanagement. Therightknowledgemanagementadoptionstrategyshouldbe put in place to develop and nurture core organizational competencies, and create intellectual capital by tapping or capturingexperiencefeedbackinformation.Ithasbeenargued thatatruevaluecreationculturecanbefoundthroughtheright combinationofhumannetworks,social,intangibleand technolo-gyassetswhereissuessuchaschangemanagement,learningand trust must be blended successfully towards the vision of knowledge-enabledvaluecreation.Therehasbeenanevolution ofknowledgemanagementovertime[25].Thedifferencesofthe evolutionareexaminedinTable1.

Incertaincircumstances(e.g.,thescarcityofspecificdata),itis not possible to move towards reliable analytical or statistical approaches [14,19] and experience feedback approach is then adopted as a more suitable alternative. However, Aamodt and Plaza[1]recommendthattheremustbeareasonablenumberof experienced cases that are sufficiently similar to be grouped togetherforpotentialapplicationofthereasoningprocessoverits underlying knowledge for use in new cases or projects. This underpinstheconceptof‘‘experiencedfeedback’’.

2.2. Knowledgemanagementthroughexperiencefeedback

The usefulness of information structuring and knowledge capitalization from experience feedback in solving engineering problems have been examined in Gardoni et al. [10,11]. The experiencefeedbackprocessconsistsofthreeessentialsteps(see

Table2):(i)thecapitalizationphase,(ii)thetreatmentphase,(iii) andtheexploitationphase.InTable2,thefirsttwostagesconsistof capitalizationandasynthesisofdifferentstagesoftreatmentin experiencefeedback.Thethirdisabouttheexploitationphaseof feedbackforfaultdiagnosissystems,safetyassessment(levelsof protection),andforecastingchangesinequipmentovertime.

Inordertogenerateknowledgefromexperience feedback,a conceptualmodel[15]incorporatingproblem-solvingtogenerate explicitknowledgefromexperiencesispresentedinFig.1.

Table1

ThreegenerationsofknowledgemanagementintheAECindustry.

GenerationsofKMcriteria Generation1:knowledgesharing Generation2:knowledge conceptualizationandnurturing

Generation3:knowledgevaluecreation

UnderpinningICT Humaninterpretableknowledge

systems(knowledgeembedded indocumentsrequiringhuman interpretation)

Semanticbasedsystems(articulated aroundtheuseofBuildingInformation Modelingorontology)

Systemsmanagingknowledgeasan asset(leveragingtheintellectualand socialcapitalofanorganization) Socio-technical

dimension

Informationandcommunication technology

Humanandorganizationalfactors Humannetworks,socialcapital, intellectualcapital,technologyassets, andchangemanagement

Lifecyclefocus Softwareapplicationfocus Disciplinefocus TotallifecycleandAECdomainfocus

Knowledgeperspective Conditionperspectiveemphasizing accesstoinformation

Knowledgeprocessperspectivefocusing onknowingandacting

Capabilityperspective:knowledgeisa capabilitywithaviewofcreatingvalue

(4)

IntheAECindustry,theappropriatedeploymentof informa-tionstructuringstrategieswithsemanticknowledgecanleadto efficient management of the key variables or factors that influence the built environment. For example, the use of semantic web and linked data to structure information for decision-makingaboutenergy,greenhousegases,waterorpoor healthhavebeenreportedin[2].Therearelessonstobelearned from experiences in diverse contextual situations. An experi-encedactorcouldexploretheknowledgeinabundancecovering differentaspectsofbuildingconstructiontoprovideimportant

workexamplesthatis effectiveandinstructiveinconstruction activities.Thisapproachinvolvestheuseofspecialist informa-tion or knowledge from available records or sources where expertswillexploitforuseintheirvariousapplicationssuchas diagnosis[6],riskanalysis[21],safetyassessmentandsecurity

[31].Infact,theconceptualizationofthesharedrepresentations significantlyinfluencesthe manner in whichthecollaborative work takes place, as evidenced by their contribution in the exchange of information and generated knowledge. Ontology provides a better way to explicitly specify concepts of any domain[13].Therefore,inordertogainabetterunderstandingof domainandtobeabletomakecomprehensiveanalysisof the experiencefeedbackprocessesintheAECdomain,an ontology-driven approach is adopted. Furthermore, ontologies provide meanstoformalize domainknowledgein away thatmakes it accessible,shareable,reusableandmachineprocessable. 3. Ontology-drivenapproachforknowledgemodelinginAEC

In this section, an ontology-driven approach for knowledge modelinginAECisproposed.Thisapproachprovidesopportunities forsustainedcontinuousimprovementandre-useofknowledgein AEC industry. In software engineering the task of rendering information toanend-userthroughtheimplementation of any computer system often involves three separate layers: the presentation, the application or business logic and data (data managementprocesses)layers(seeFig.2).Thisisoftencalledthe hybrid system, client-server architecture or 3-tier architecture.

Table2

Experiencefeedbackprocessinlinewithcontinuousimprovement[24]. Phasesof experience feedback Stagesof problem solving

Definitionofthestage

Capitalization Problem Identificationofproblem

Treatment Reflection Analysisofproblem

Origins Investigationofcauses

Guide Searchforpossiblesolutions

Reaction Selectionofthemost

adequatesolution

Exploitation Execution Implementationofthesolution

Monitoring Monitoringofactionsand

verificationoftheeffectiveness Standardization Generalizationtopreventthe

recurrenceofsimilarproblems

Fig.1.Experiencemodelincorporatingproblem-solvingtechnique(KamsuFoguemetal.[15]).

(5)

Access to the presentation layer is often through the useof a graphicaluserinterface(GUI)orawebbrowser.Asarguedby[3], theimplementationofthe3-tiermodelinitsfullgeneralityandthe invocationofaccessviatheinternetorwebbrowseriscalledan N-tier model. The N-tier model has the following advantages: adaptability,encapsulation,re-useandquality.

The implementation will be performed according to each technologyinvolvedinagivenlayer.Inordertodemonstratehow eachlayerfeedstheother,thedatalayerimplementationwillbe conductedbeforetheapplicationlayer.Itisimportanttonotethat withinthescopeofthisstudy,thefocuswillonlybeonthedata and business logic layers. The development of a user-friendly interface(i.e.thepresentationlayer)isreservedaspartoffuture study.

Withregardstothedatalayer,themaintechnologyinitisthe CoGuigraphicalknowledgerepresentationmodelwhichcanstore informationaboutanydomainofdiscourse[29].Thedatacaptured inCoGuiisfromexternaldocumentsandstoragesystemswhich could be non-structured, semi-structured or structured. Once editedintoCoGui,theCoGuiknowledgemodelcanbeconverted intodifferentfile/datamodelformats(e.g.,ResourceDescription Framework(RDF))forotherusessuchasinapplication develop-ment.Themodelcanalsobeenrichedwithconceptualgraphrules andconstraintstocapturereallifesituations[17,35–43].Therules andconstraintsarebasedonConceptsTypes,RelationshipTypes andIndividuals.Withregardstothebusinesslogiclayer,itsmain purpose is toextract relevant partsfromthe CoGuiexperience feedback knowledge base,stored in theform of a CogXML file format. This data is obtained from the data layer. It contains the conceptualgraphsbase‘‘CG base’’that communicateswith the ‘‘Conceptual Graph Manager’’.The ‘‘Projection Engine’’is a reasoning engine based on the semantic mapping between conceptualgraphsinvolvingbothadjacentconceptsandrelations. The constraint checking is a procedure for verification of the essential requirements and expectations in the CG base. The ‘‘InferenceEngine’’isareasoninganddeductiveprocessapplying rules(iftheirconditionsaretrue)totheavailablefactsinorderto generatenewconclusions.Infact,therulesareappliedonlywhen requiredconditionsareincompliancewithconsideredfactsand/or available hypotheses. The crucial aspect of the architecture presentedinFig.2istheontologydevelopmentusingCoGui.All whatisrequiredtobeundertakeninthedataandbusinesslogic layers canbeundertakenwithintheCoGuienvironment.Inthe ensuing sections, the ontology development procedures using CoGuiwillbeexamined.

3.1. Graph-basedontologydevelopmenttechnique

To develop theontology, a graph-based technique hasbeen usedtorepresentknowledgeaboutthedomainofinterest.This techniqueenablesreal-timechangestobemadetofindfasterand clearersolutionsfordifferentapplications[8].Foranyknowledge representation,itisimportanttoestablishtherequirementstobe fulfilledbytheapplicationforwhichtheknowledgemodelisbeing developedfor.Tothisend,thefollowingrequirementshavebeen set:

Awiderapplicationfieldtoallowthesatisfactionofcertainneeds ofexpandingthegroundinwhichknowledgecanbeshared.This constitutes thefirstbigstep forhavinga properly positioned experience feedback process in a continuous improvement environment.

A formal conceptualization of a collaborative work system shouldbeconsideredforpurposesofsharedrepresentationsin supporting the process of collective knowledge creation, evaluationandlearningfromexperiences.

A simple and clear procedure to manage and exploit the appropriateinformation.Itishighlyadvisabletogofortheless crowdedand themostexplicit wayof datarepresention and visualization

Afeasibleimplementationofthesolutiontoreality:Thisimplies thatthesolutionshouldbringmorebenefitsthanproblems.Itis knownthatthefullexploitationrequiresaperiodoffinetuning intermsofuseandoperationalapplication.

The AEC industry is noted for generating huge amount of informationfromconstructionprojects.Forthisinformationtobe effectively shared, a well-organized method is required to minimizeambiguity.Hence,therationaleforadoptinganontology approach to disambiguate inconsistencies and facilitate under-standingandreasoning.Asexplainedearlier,anontologyformally representsknowledgeasasetofconceptswithinadomain,andthe relationshipsbetweenpairsofconcepts.Itcanbeusedtomodela domainandsupportreasoningaboutentities.Fromthis perspec-tive,ontologyprovidesaviableandefficienttechniqueofsharing experiencefeedbackknowledge.Asarguedin[30],oneofthebest waystoemployontologytocapturetheaboverequirementsisto adopt conceptual graph representation. To this end, CoGui, a graphicalanddevelopmentontologyeditorisusedinmodelinga realworldapplicationofexperiencefeedbackintheconstruction domain.

3.2. Knowledgemodellingwithconceptualgraphs

ConceptualGraphsuserinterface(CoGui1₎_is_a_free_graph-based

visual tool, developed in Java, for building conceptual graph knowledgebasesrepresentedinCogXMLformat,compatiblewith Cogitant[4].CoGuimakesitpossibletocreate,editandcontrolthe structureandcontent,ofaknowledgebase.Theknowledgebase can be serialized in the XML format called CogXML. Such a knowledgebaseiscomposedofanontologicalknowledge,afactual knowledge represented by a set of conceptual graphs and proceduralknowledgedescribed byconceptualgraphrules(see

Fig.3).Theconceptualgraphcanbesimpleornestedforembedded descriptionsofknowledgerepresentation.

3.2.1. Ontologyrepresentation

Accordingto[5],thebasiccomponentsofanontologyis the vocabulary,whichconsistsof:

A hierarchy of concept types (also called concepts), which representsthekindsofentitiesintheapplicationdomain.These concepts could be defined to be disjoint or to have an intersection;

A hierarchy of relation types (also called relations), which representsthekindsofrelationshipsbetweenentities,andmay haveanynumberofarguments;

Asetofindividuals(orindividualnames)andgraphsassociated with individuals, which represent knowledge about these individuals;

Asetofconceptualgraphrulestakestheformofanimplication between an antecedent(body)and a consequent(head). The intendedmeaningisreadas:whenevertheconditionsspecified intheantecedentholds,thentheconditionsintheconsequent mustalsohold.

3.2.2. Conceptualgraphsformalism

ConceptualgraphswereintroducedbySowaasadiagrammatic systemoflogicwiththepurpose‘‘toexpressmeaninginaform

(6)

that is logically precise, human readable and computationally tractable’’[28].Conceptualgraphsencodeknowledgeas graphs andthuscanbevisualizedinanaturalway[30]:

Thevocabulary,whichcanbeseenasabasicontology,ismadeof hierarchiesofconceptsandrelations.

Allotherkindsofknowledgearebasedontherepresentationof entitiesandtheirrelationships.Thisrepresentationisencoded by a labeled graph, with two kinds of nodes, respectively correspondingtoentitiesandrelations.Edgeslinkanentitynode toarelationnode.

Aruleexpressesimplicitknowledgeoftheform:‘‘ifhypothesis, thenconclusion’’,wherethehypothesisandconclusionareboth basicgraphs.Usingsucharuleconsistsofaddingtheconclusion graph(tosomefact)whenthehypothesisgraphispresent.There isaone-to-onecorrespondencebetweensomeconceptnodesof thehypothesiswithconceptnodesoftheconclusion.Twonodes incorrespondencerefertothesameentity.Thesenodesaresaid tobeconnectionnodes.Theknowledgeencodedinrulescanbe madeexplicitbyapplyingtherulestospecifiedfacts.

Having discussed the graph-based ontology development processinSection3,inSection4,thestepswillnowbeapplied toacasestudy.

4. AnillustrativemodelingcaseintheFrenchcontext AFrench AECcompanylocatedin thesouthwestregion has beenusedasacasestudytoillustrateknowledgemodelingofthe experiencefeedbackprocess.Thiscompanyisactiveintwosectors: publicworksandbuilding.Themainactivitiesoftenundertakenby thecompanyarepublicworks(e.g.,roads,railroads,bridges,water supply,sewageandelectricalgrid)andbuildings(e.g.,universities andhospitals).Thisstudyfocusedonfacilitatingworksasdefinedby the UKNew Rules of Measurement (NRM)[26].The facilitating workswererelatedtosixtramwayprojectsbeingconstructedby companyX(fordataprotectionpurposes,Xisassumedtobethe nameofthecompany)intheSoutheastregionofFrance.Facilitating works include specialist works which normally need to be completedbeforeanybuildingworkscancommence(e.g.,major demolition works, soil stabilization works and/or temporary diversionofmainsdrainage).Thekeychallengesassociatedwith construction of tramway platformswere the qualityand safety needs,theregulatoryconstraintsandminimumtrafficdisruptions. The rationale for choosing the UK NRM is because of its well-establishedstructureandpopularityofitsuseinothercountries.It

providesdefinitionsfordifferentworkelementsinaconstruction projectthanmoststandardmeasurementmethods.Byadoptingthe NRMstandards,itiseasiertomapconceptselicitedfromfeedback processes.Intheadoptedmodelingapproach,anindividual(e.g., experience feedback manager) in the company responsible for experience feedback treatment undertook the following tasks discussedintheensuingparagraphs.

Firstly, information fromcompany’s project are gathered or elicited.Thiscanbeachievedin two ways[33]. Theexperience feedbackmanagercompletesarecordofinformationgatheredfrom thetreatmentofexperiencedcases,andthenconductsinterviews withkeyactorsofprojectshavingpotentialinterestinglessonsthat canbelearned.Basedonthese,thefeedbackmanagerestablished records of collected information in order to generate useful knowledge.Secondly,thestructuredorsemi-structuredformsof experiencefeedbackactivitiespreparedbythefeedbackmanager whichcanalsobeobtainedandpreparedbyanyactorsthrough interviewsaresavedinadatabase.Thisisstoredforpurposesof knowledgereuseinAEC[20].Aftercompletingthiscollectionof information,itisanalyzedtodeterminethetype,richnessandits usefulness. The identification of different types of information dependsontheassociatedknowledgeandalsoontheirimportance. Examples of categories includeinformation gathered associated withonsitepracticalproblems(e.g.,lackofsafetyonsite)andsimple information that can be useful afterwards (e.g., use of a new material).Withregardstothetramwayprojects,someactionswere undertakentoelicitknowledgeaboutexperiencefeedback process-es.TheactionsimplementedaresummarizedinTable3.

Ontheonehand,intheengineeringphaseofaconstructionofthe tramwayprojectsthecapitalizationconcernsdatacollectionfrom specific projects related to design phase (new materials, used efficiency coefficients, particular production mechanisms.). The treatmentisaboutthedifferentwaysusedtoaddressthefacilitating work challenges and the search for reasonable solutions. The exploitationdefinesthemeanswithinforinformationretrievaland thedisseminationoftheinterestingresultsoflessonslearnt.Onthe otherhand,intheengineeringphaseofaconstructionprojectthe capitalizationisfortheexecutionandcoordinationofthestudiesby thearchitecturalandengineeringteamandfortheactualworks.The treatmentconcernstheactionsundertakeninfindingsolutionsto recurrentproblemsandmorefeasiblealternatives.Theexploitation offeedbackisachievedthroughmeetingswiththeaimofpreventing problems,ortodeliversolutions.

Uponcollecting theinformationasdescribed inTable3,the followingkey ontologicalconceptselicited weredevelopedand willbediscussedintheensuingsections.

(7)

4.1.1. Concepttypes

Basedonthedefinitionofthefacilitatingworks[26],concepts orclasseswiththeirrespectivesub-conceptswereabstractedand modeledinCoGui(seeFig.4).

Based on Fig. 4, the first levels in the hierarchy are Toxic/ Hazardous/contaminated material treatment, Major demolition

works, Temporary support to adjacent structures, Specialist groundworks,TemporarydiversionworksandExtra-ordinarysite investigation works.These toplevel conceptsalso contain sub-concepts.Somerelationsmaybeestablishedbetweentheconcepts andusedforthemodelingoffactualandproceduralknowledgein conceptual graphs. This can facilitate automated reasoning in experiencefeedbackprocesses.

4.1.2. Relationtypes

Fig.5depictstherelationships(Logicaloperators,Comparison operators, Usual relations and Temporal relations) between concepts, including their respective branches created. Logical operators(suchasAND,NOT,ORandXOR)areusedincombining twoormorerelationalexpressionsintoasingleexpressionthat evaluatestoeithertrueorfalse.Comparisonoperators(Equal,Less thanandGreaterthan)canbeusedtocomparetwoconceptswith thelogicaltrueandfalseresults.Usualrelations(suchasAgent, Attribute and Object) refer tothe construction of sentences in termsofsubject,verbandcomplementinthecommonlanguage withactiveandpassivecomponents.Temporalrelations(suchas After,beforeandduring)areusedforthetemporaldescriptionsof events.

4.1.3. Factsandruleapplications

After creating the concept and relation types, the relevant constraintsandrulesarecreatedtofacilitateunderstanding,only rules related to four sub-elements of the third level will be examined. The elements are ‘‘Toxic or hazardous material removal’’,‘‘Demolitionworks’’,‘‘Soilstabilizationmeasures’’and ‘‘Archaeologicalinvestigation’’.Toxicorhazardousmaterialremoval

is a concept describing the removal, employing special safety measures,oftoxic orhazardousmaterialprior todemolitionor refurbishment works. Fig. 6 represents the modeling of an associated rulein the conceptualgraph formalism. Therule in

Fig.6means,ifacertainprojecthascomponents(object)witha toxicityvaluegreaterthanagiventhresholdlevelthentheobject shouldbedetoxified.

Logicalexpression:

9

x

9

y

9

z(Project(x)^ Toxicity(y)^ Upper Threshold (z)^ Attribute (x, y)^ Superior (y, z)!

9

_u _(Project

(x)^ Detoxification(u)^ Object(u,x))

Soilstabilizationmeasuresisaconceptdescribingthe stabiliza-tionorimprovementofsoilbearingcapacityorslipresistanceof existinggroundtofacilitateconstructionbyinjectingorotherwise

Fig.4.Concepttypesoffacilitatingworks. Table3

Experiencefeedbackactivitiesconductedinthecasestudy. Executionphases ofaconstruction project. Typeofexperience feedback Activitiesconducted

Engineering Capitalization Datacollectionfromspecificprojectsrelatedtodesignphase(newmaterials,usedefficiencycoefficients,particular productionmechanisms.)

Interviewwithsiteworkers,foremen,sitesupervisorsandconstructionmanagers. Treatment Thepersonisseekingalonesomesolutionstotheproblemsduringitsworkingtime.

Talkswithothersaffectedbytheproblemforcollaborativeproblemsolving. Exploitation Ascertainedandrecordedsolution.

Solutiontransmitted(orally,bymail,byphone...)toindividualsinvolvedinthestudyphase.

Construction Capitalization Collecteddata:

Passedfromtheforemantositesupervisors(securityissues,problemsonamachine,amaterialused,bad equipment...).

Amendmentsmadeduringquality/safetymeetings.

Executionandcoordinationofthestudiesbythearchitecturalandengineeringteamandfortheactualworks. Treatment Theactorsaretryingtofindsolutionstorecurrentproblems

Exploitation Memorandum,diagrams,oranyotherpracticaldocuments. SafetyDataSheet.

Meetingstopreventexistingproblems,ortodeliversolutions.

(8)

introducing stabilising materials, by power vibration, by soil nailingorbygroundanchors.Fig.7representsthemodelingofan associatedruleintheconceptualgraphformalism.TheruleinFig.7

meansthatif thevalueof thesoilphysical stabilitywherethe

projectisbeingundertakenislessthanagiventhresholdvalue,a soilstabilizationmeasureshouldbeperformedwhichconsistsof improving the soil bearing capacity and/or slip resistance of project.

Fig.5.Relationtypes.

Fig.6.Arulemodelingasituationforthedetoxification.

(9)

Logicalexpression:

9

x

9

y

9

z(Project(x)^ SoilPhysicalstability (y)^ LowerThreshold(z)^ Attribute(x,y)^ Inferior(y,z))!

9

u

9

v

9

_w_(Project_(x)^ Soilstabilizationmeasures(u)^ Bearingcapacity (v)^ Slipresistance(w)^ Object(x,u)^ Element(u,v)^ OR(v,w)). Physical archaeological investigation works are site based excavationworksin searchofartefacts andthelikecarriedout bythemaincontractorundertheguidanceandinstructionofan archaeologist.Fig.8representsthemodelingofanassociatedrule intheconceptualgraphformalism.TheruleinFig.8meansthatifa projectislocatedonanarchaeologicalgrounds,thenexcavations works and the search of archaeological artefacts should be undertaken.

Logicalexpression::

9

x

9

y(Project(x)^ Archaeological investi-gation(y)^ Element(x,y)!

9

u

9

v(Project(x)^ Excavationworks (u)^ Searchofartefacts(v)^ Element(x,u)^ Object(u,v))

Demolitionworksconcernthetakingdowntogroundleveland removing complete buildings/structures or parts of buildings/ structures, includingservices,fittingsand finishes.Fig.9 repre-sentsthemodelingofanassociatedruleintheconceptualgraph formalism.TheruleinFig.9meansthatifmajordemolitionworks are to beundertaken on a project’ssite, all services’ networks shouldbedisconnected.

Logicalexpression:

9

x

9

y(Project(x)^ Majordemolitionworks (y)^ Element (x, y)!

9

u (Project (x)^ Network disconnection (u)^ Element (u, v)^ Mains Services ({water, gas, electrici-ty})^ Object(u,{water,gas,electricity}))

BasedonfactspresentedinFigs.6and7,anillustrationofthe general principle regarding the application of any rule will be examined.ThisispresentedinFig.10,wheretheselectionofafact activatesanapplicablerule.Theapplicationofthetwoprevious factsandrulesconcernsthefacilitatingworksfortheconstruction ofatramwaymanagementplatformintheoutskirtsofToulouse.

Theplatformprovidesservicessupportingthecoordinationand the monitoring of tramway vehicles serving the community includingtheremoteparts.

Therules6–9areimplementedinCoGui.Afterthefactandrules areselected,thesimulationislaunched.Thesimulationresultswill serveasabasistohighlightissuesthatneedtobeaddressed.

Asanexample,therightofFig.10istheoutputoftheexecution ofthefactsontheleftofthesamerule.Itisimportanttonotethat thevisibilityofthesimulatedresultsinCoGuiwasverypoor,and couldnotbereadwhenthescreenwassnapped.Consequently,the outputontherightofFig.10iscolouredgreenjusttoindicatethat itisanactualoutputfromCoGui.

5. Discussions

Inthecasestudyexamined,akeyobservationthatcanbemade abouttheprocedureisthefactthatchallengesassociatedwiththe facilitatingworkscanbemodeled usingconceptualgraphsand enriched with rules to facilitate reasoning. By developing an ontologyknowledgebasethatincludesfeedbackprocesses,it is possible to re-use in future in solving a large majority of problematicsituations.Anexampleofapracticalproblemsolved isthedecisiontostabilizethesoilifthesoilbearingcapacityis lower than a given threshold value (Fig. 7). Reasoning in the knowledgebasedisfacilitatedbyboththeProjectionandInference enginesinCoGui.Inadditiontostrengthsoftheseengines,CoGui conceptualgraph-basedstructuremakesiteasytoorganizeand sortinformationthatcanbeusedincorporationsforcontinuous improvementinconstructionprojects.InFrance,theapplicationof experiencefeedbackinimprovingorganizations’efficiencyismore commoninthefieldofindustrialengineeringthaninAECsector. The main reason is because of the automation of production systemsin industrial engineeringcompounded bythefact that

Fig.8.Arulemodelingasituationforthearchaeologicalinvestigation.

Fig.9.Arulemodelingasituationforthemajordemolitionworks.

(10)

industrialproductsoftenhavesimilarcharacteristics.Thus,itisnot uncommontofindexperiencefeedbackapplicationsinindustrial repairs and maintenance management ([16], [18]; Potes [23]). While experience feedback process can easily be applied in industrial production/engineering, because most of its tasks/ productsareoftenrepetitiveandstandardized,itsapplicationin AECisverychallenging.ThisisbecauseeachAECprojectisoften unique in nature. With regards to renovation/refurbishment projects,thechallengeisevenenormousaseachexistingbuilding structure has different characteristics and different levels of defectsrequiringdifferentappraisalstrategy orsurveying tech-niquesbeforerecommendationforrepairsandmaintenance.That notwithstanding,andgiventheknowledgeintensivenatureofAEC projects,developinganontologyknowledgebaseaboutexperience feedbackprocessesprovidesanopportunitytostructure knowl-edgethatcanbere-usedinsolvingproblemsondifferentprojects.

6. Conclusionandrelatedworks

Therationaleforthisstudystemmedfromthefactthatknowledge isanimportantassetofanorganizationincludingAECcompanies; manyAECcompaniesareyettotaponthefullpotentialofknowledge fromalreadyexecutedprojects.Companiesareawareofsituationsin theirorganizationsinwhichcostlyerrorshavebeenmadebecause knowledgewasnotavailablewhenandwhereitwasneededand because employees did not know how to interpret or use the informationavailabletothem.Thesesituationsaremainlyduetothe project-basednatureoftheconstructionindustryandthefactthat knowledgeisembeddedinsocialrelations.Nevertheless,forthosebig companieswhomight becarryingout internationalprojects,strategies to mobilize knowledge is critically important. In construction organizations, time is often associated with the need to deliver projects according to schedule. Many construction organizations believethattheirorganizationalstructurerequiresheavyprocedures andlongprocessingtimestoexploitdomainknowledge.Employees maybewillingtoshareknowledge,butthepressuretodeliverprojects undertightscheduleandbudget makesthemhaveverylowappetiteto takeonadditionalresponsibilityforknowledgemanagement activi-ties.Inthismanuscript,itwasrevealedthatusefulknowledgeonAEC projectsandbusinessesororganizationscanbehelpfulinresolving persistent, complex and workload problems. Eliciting experience feedbackknowledgeiscoretore-usingexistingknowledgethatcan improveorganizationalefficiencyinproduct(e.g.,buildings)delivery. Thisisenshrinedinthemethodologyforknowledgecapitalization fromsemi-structuredinformationandunstructuredinformationin industrialengineeringdomain[10,11].Graphicalconceptualmodeling isvery powerful asit allowslessons learned to be capturedand representedtextuallyandgraphically,facilitatingvisualization.This improves communication to users who may be interested in consuming knowledge about feedback from previous projects

[27]. In this study, we adopted the N-tiermodel to develop the ontologyknowledgebasecontainingexperiencefeedbackknowledge fromtramwayprojectsconstructedintheSoutheastregionofFrance. Thefocusedofthepaperwasonthedataandbusinesslogiclayers.As partoffuturestudy,thedevelopmentofauser-friendlyinterfacewill beinvestigated.Itisalsoimportanttonotethat,althoughtheontology developedisbased ontramway projectsinSoutheast France,the concepts andrules are genericand can be re-used on any other tramwayprojectsinanycountry.Suchamoveshouldonlyfocuson facilitatingworks.

Acknowledgement

Thiswork hasbeen undertakenas part ofthe collaborative research funded through two organisations whosegoals are to

encourage interdisciplinary research work. Firstly, the authors were awarded a grant by the Ecole Nationale d’Inge´nieurs de Tarbesthrough the‘‘Projet Soutien a´ la Mobilite´ Internationale (SMI) 2014’’ programme. Secondly, the authors were awarded anothergrantbytheOxfordBrookes UniversitythroughCentral ResearchFundingScheme.Theauthorsgratefullyacknowledgethe financialsupportreceived.

References

[1]A.Aamodt,E.Plaza,Case-basedreasoning:foundationalissues,methodological variations, andsystem approaches, Artificial Intelligence Communications7 (1994)39–59.

[2]F.H.Abanda,J.H.M.Tah,R.Keivani,TrendsinbuiltenvironmentsemanticWeb applications:wherearewetoday?ExpertSystemswithApplications40(2013) 5563–5577.

[3]G.Alonso,F.Casati,H.Kuno,V.Machiraju,WebServicesConcepts,Architectures andApplications, SpringerVerlag,Berlin,Germany,2003.

[4]J.F.Baget,M.Chein,M.Croitoru,A.Gutierrez,M.Lecle`re,M.L.Mugnier,Logical, graphbasedknowledgerepresentationwithCoGui,in:Actesdel’atelierGAOC (GraphesetAppariementd’ObjetsComplexesenconjonctionavecEGC’10),2010, 15–25.

[5]M.Chein,M.Mugnier,Graph-basedKnowledgeRepresentation:Computational FoundationsofConceptualGraphs, Springer,London,UK,2009.

[6]C.-Y.Chang,M.-D.Tsai,Knowledge-basednavigationsystemforbuildinghealth diagnosis,AdvancedEngineeringInformatics27(2013)246–260.

[7]A.S.Choo,K.W.Linderman,R.G.Schroeder,Methodandcontextperspectiveson learningandknowledgecreationinqualitymanagement,JournalofOperations andManagement25(2007)918–931.

[8]R.Dieng,O.Corby,Conceptualgraphsforsemanticwebapplications,in:F.Dau, M.L.Mugnier,G.Stumme(Eds.),Proceedingsofthe13thInternationalConference onConceptualStructures(ICCS’2005),Springer-Verlag,Kassel,Germany,2005, pp.19–50(July17-23,2005,LNAI3596).

[9]N.Forcada,A.Fuertes,M.Gangolells,M.Casals,M.Macarulla,Knowledge man-agement perceptionsinconstruction and design companies,Automation in Construction29(2013)83–91.

[10]M.Gardoni,M.Spadoni,F.Vernadat,Harnessingnon-structuredinformationand knowledgeandknow-howcapitalisationinintegratedengineering:casestudyat AerospatialeMatra,ConcurrentEngineering8(4)(2000)281–296.

[11]M.Gardoni,C.Frank,F.Vernadat,Knowledgecapitalisationbasedontextualand graphicalsemi-structuredandnon-structuredinformation:casestudyinan industrialresearchcentreatEADS,ComputersinIndustry56(1)(2005)55–69.

[12]J.Goulding, W.Nadim,P.Petridis,A.Alshawi, Constructionindustryoffsite production:avirtualreality interactivetrainingenvironmentprototype, Ad-vancedEngineeringInformatics26(2012)103–116.

[13]T.R.Gruber,Atranslationapproachtoportableontologyspecifications? Knowl-edgeAcquisition5(2)(1993)199–220.

[14]J.Juha´sova´,R.Sofronie,R.Bairra˜o,Stonemasonryinhistoricalbuildings—Waysto increasetheirresistanceanddurability,EngineeringStructures30(8)(2008) 2194–2205.

[15]B.Kamsu-Foguem,T.Coudert,C.Be´ler,L.Geneste,Knowledgeformalizationin experiencefeedbackprocesses:anontology-basedapproach,Computersin In-dustry59(2008)694–710.

[16]B. Kamsu-Foguem, Knowledge-based support innon-destructive testing for healthmonitoringofaircraftstructures,AdvancedEngineeringInformatics26 (2012)859–869.

[17]B.Kamsu-Foguem,G.Diallo,C.Foguem,Conceptualgraph-basedknowledge representationforsupportingreasoninginAfricantraditionalmedicine, Engi-neeringApplicationsofArtificialIntelligence26(2013)1348–1365.

[18]B.Kamsu-Foguem,D.Noyes,Graph-basedreasoningincollaborative knowl-edgemanagementforindustrialmaintenance,ComputersinIndustry64 (2013)998–1013.

[19]K.Kovler,N.Roussel,Propertiesoffreshandhardenedconcrete,Cementand ConcreteResearch41(2011)775–792.

[20]K.Y.Lin,S.H.Hsieh,H.P.Tserng,K.W.Chou,H.T.Lin,C.P.Huang,K.F.Tzeng, Enablingthecreationofdomain-specificreferencecollectionstosupport text-basedinformationretrievalexperimentsinthearchitecture,engineeringand constructionindustries,AdvancedEngineeringInformatics22(2008)350–361.

[21]H.-L.Liu,Y.-L.Tsai,Afuzzyriskassessmentapproachforoccupationalhazardsin theconstructionindustry,SafetyScience50(2012)1067–1078.

[22]A.Moum,C.Koch,T.I.Haugen,Whatdidyoulearnfrompracticetoday?Exploring experiencesfromaDanishR&Deffortindigitalconstruction,Advanced Engi-neeringInformatics23(2009)229–242.

[23]P.A.P.Ruiz,B.Kamsu-Foguem,D.Noyes,Knowledgereuseintegratingthe col-laborationfromexpertsinindustrialmaintenancemanagement, Knowledge-BasedSystems50(2013)171–186.

[24]H.Rakoto,P.Clermont,L.Geneste,Elaborationandexploitationoflessonslearned, 93,IFIP–TheInternationalFederationforInformationProcessing,Montre´al,Que´bec, Canada,2002,pp.297–300.

[25]Y.Rezgui,C.J.Hopfe,C.Vorakulpipat,Generationsofknowledgemanagementin thearchitecture,engineeringandconstructionindustry:anevolutionary per-spective,AdvancedEngineeringInformatics24(2010)219–228.

(11)

[26]RICS(2012)NRM1.Orderofcostestimatingandcostplanningforcapitalbuilding works.UK,TheRoyalInstitutionofCharteredSurveyors.

[27]J.D. Silvestre,J. deBrito,Ceramictiling inbuildingfac¸ades:inspectionand pathologicalcharacterizationusinganexpertsystem,ConstructionandBuilding Materials25(2011)1560–1571.

[28]J.F.Sowa,Conceptualgraphsforadatabaseinterface,IBMJournalofResearchand Development20(1976)336–375.

[29]J.F.Sowa,ConceptualStructures:InformationProcessinginMindandMachine, Addison-Wesley,Reading,MA,1984.

[30]J.F.Sowa,KnowledgeRepresentation:Logical,Philosophical,andComputational Foundations, Brooks/ColePublishingCo,PacificGrove,CA,2000.

[31]D.Theodossopoulos,B.Sinha,Areviewofanalyticalmethodsinthecurrent designprocessesandassessmentofperformanceofmasonrystructures, Con-structionandBuildingMaterials41(2013)990–1001.

[32]W.Tizani,M.J.Mawdesley,Advancesandchallengesincomputinginciviland buildingengineering,AdvancedEngineeringInformatics25(2011)569–572.

[33]E.Wenger,R.McDermott,W.M.Snyder,CultivatingCommunitiesofPractice, HarvardBusinessSchoolPress,Boston,2002.

[34]X.Zhang,X.Mao,S.M.AbouRizk,Developingaknowledgemanagementsystem forimprovedvalueengineeringpracticesintheconstructionindustry, Automa-tioninConstruction18(2009)777–789.

[35]B. Kamsu-Foguem, Ontological view intelemedicine, EuropeanResearch in Telemedicine/LaRechercheEurope´enneenTe´le´me´decine3(2)(2014)67–76.

[36]B.Kamsu-Foguem,G.Tchuente´-Foguem,C.Foguem,Verifyingamedicalprotocol withtemporalgraphs:Thecaseofanosocomialdisease,JournalofCriticalCare29 (4)(2014)690.e1–690.e9.

[37]B.Kamsu-Foguem,G.Tchuente´-Foguem,C.Foguem,Conceptualgraphoperations for formal visualreasoninginthemedical domain,IRBM-Innovationand ResearchinBioMedicalEngineering35(5)(2014)262–270.

[38]B.Kamsu-Foguem,D.Noyes,Graph-basedreasoningincollaborativeknowledge managementforindustrialmaintenance,ComputersinIndustry64(8)(2013) 998–1013.

[39]B.Kamsu-Foguem,G.Tchuente´-Foguem,C.Foguem,Usingconceptualgraphsfor clinicalguidelinesrepresentationandknowledgevisualization,Information Sys-temsFrontiers16(4)(2014)571–589.

[40]B. Kamsu-Foguem, A. Flammang,Knowledge description for the suitability requirementsofdifferentgeographicalregionsforgrowingwine,LandUsePolicy 38(2014)719–731.

[41]B.Kamsu-Foguem,V.Chapurlat,Requirementsmodellingandformalanalysis usinggraphoperations,InternationalJournalofProductionResearch44(17) (2006)3451–3470.

[42]P.A.P.Ruiz,B.Kamsu-Foguem,D.Noyes,Knowledgereuseintegratingthe col-laborationfromexpertsinindustrialmaintenancemanagement, Knowledge-BasedSystems50(2013)171–186.

[43]P.A.P.Ruiz,B.Kamsu-Foguem,B.Grabot,Generatingknowledgeinmaintenance fromExperienceFeedback,Knowledge-BasedSystems68(2014)4–20.

Dr.BernardKamsu-Foguem iscurrently atenured AssociateProfessorattheNationalEngineeringSchool ofTarbes(ENIT)ofNationalPolytechnicInstituteof Toulouse (INPT) and undertakes research in the ProductionEngineeringLaboratory(LGP)ofENIT-INPT, a research entity (EA1905) of the University of Toulouse.Heisavisitingscholarinsomerenowned universities: United Kingdom (e.g. Oxford Brookes University, Oxford), Finland (e.g. Aalto University, Helsinki University of Technology, VTT Technical Research Centre of Tampere, University of Oulu, A˚boakademiofTurkuUniversity).Heisamemberof

InternationalEditorialReviewerBoardofArtificialIntelligenceResearchanda reviewerforalargenumberofinternationalscientificjournalssuchas Knowledge-BasedSystems,EngineeringApplicationsofArtificialIntelligence,International JournalofProductionResearch,JournalofIntelligentManufacturing,Computer Methods and Programs in Biomedicine, Computers in Biology and Medicine, InteractingwithComputers, SensorsandKnowledge ManagementResearch& Practice. He has a Master’s in Operational Research, Combinatorics and Optimization(2000)fromNationalPolytechnicInstituteofGrenoble,andaPhD inComputerScienceandAutomation(2004)fromtheUniversityofMontpellier2. Based on his research activities and outputs including 30 peer-reviewed publicationsamongstothers,heachievedthe‘‘accreditationtosuperviseresearch’’ status(abbreviatedHDR)fromINPTin2013.ThisstatusallowsDr.Kamsu-Foguem tobeabletoindependentlysupervisePhDstudents.Hisrecentresearchinterest focusesonmethodologiesandmodellingrelatedtoknowledgerepresentationand reasoningmethods ofArtificial Intelligence thatsupportknowledge-intensive activities with interdisciplinary perspectives in different applications (e.g. industrialengineering,constructionindustryandhealthcare).Heisamember of the group: e-Health of InterOP-VLab (International Virtual Laboratoryfor EnterpriseInteroperability). Theaim ofInterOP-VLab is tolink together ina network researchers and research institutions and industry representatives, engagedindevelopingapproachesandintegrativesolutionstoconnect heteroge-neousindustrialsystems,publicadministrationsororganizations.InterOP-VLab wasfoundedin2007andisthecontinuationoftheINTEROPNetworkofExcellence (Interoperabilityresearchfornetworkedenterpriseapplicationsandsoftware),a researchinitiativeoftheEuropeanUnionfoundedearly2000s,whichdevelopedthe ModelDrivenInteroperabilityFramework.

Dr.FonbeyinHenryAbandaisaSeniorLecturerinthe Department ofRealEstate &ConstructioninOxford BrookesUniversity,UK.Henryisavisitingscholaratthe EcoleNationaled’Inge´nieursdeTarbes,anengineering institutioninInstitutNationalPolytechniquede Tou-louse,France.HeisaCharteredEngineerwiththeUK Institution of Engineering and Technology. He has achievedtheRenewableEnergyExpertstatusfromthe EuropeanEnergyCentrebasedinScotland.HehasaBSc (Hons)andDipl.-Ing.inMathematics/PhysicsandCivil Engineeringrespectively.Afterobtaininghisdegreein CivilEngineeringin2003,HenryworkedasaProject Engineeronprojects fundedby thegovernmentsof CameroonandJapan.HenryobtainedhisPhDfromtheFacultyofTechnology,Design &EnvironmentatOxfordBrookesUniversityintheUKin2011.HisPhDinvestigated theextenttowhichSemanticWebtechnologiescanbeusedinmanagingsustainable buildingtechnologyknowledgeforuseindifferentbuildingprojects.Asamain contribution,thestudyculminatedinthedevelopmentofaphotovoltaicdecision supportsystemcalledPhotoVoltaicTechnologyOntologySystem(PV-TONS).The systemprovidesameansofdesigningandselectingphotovoltaic systemsand componentsforuseinbuildingprojects.HenryiscurrentlyextendinghisPhDstudyto includeothersustainablebuildingtechnologiessuchaswindturbines,geothermal andcombinedheatandpowersystems.DuringandafterhisPhDstudy,Henryhas contributedtowritingresearchproposalsforsecuringresearchgrants/fundingfora numberofresearchcouncilsandorganisationsincludingtheEngineering&Physical SciencesResearchCouncil,theInternationalLabourOrganisation,OxfordBrookes ReinventionCentre,andPersimmonHomes-UK.Inaddition,Henryhasworkedasa ChapterScienceAssistantattheIntergovernmentalPanelonClimateChangewhere hecontributedtothewritingoftheFifthAssessmentReportonclimatechange mitigationrecentlypublishedin2014.