Abstract
Abstract Maintenance procedures for complex engineering systems are increasingly determined by predictive algorithms based on historic data, experience and knowledge. Such data and knowledge is accompanied by varying degrees of uncertainty which impact equipment availability, turnaround time and unforeseen costs throughout the system life cycle. Once quantified, these uncertainties call for robust forecasting to facilitate dependable maintenance costing and ensure equipment availability. This paper builds on the theory of spatial geometry as a methodology to forecast uncertainty where available data is insufficient for the application of traditional statistical analysis. To ensure continuous forecast accuracy, a conceptual dynamic multistep prediction model is presented applying spatial geometry with long-short term memory (LSTM) neural networks. Based in MATLAB, this deep learning model predicts uncertainty for the in-service life of a given system. The further into the future the model predicts, the lower the confidence in the uncertainty prediction. Forecasts are therefore also made for a single time step ahead. When this single step is reached in real time, the next step is forecast and used to update the long range prediction. The uncertainty here is contributed by an aggregation of quantitative data and qualitative, subjective expert opinions and additional traits such as environmental conditions. It is therefore beneficial to indicate which of these factors prompts the greatest impact on the aggregated uncertainty for each forecast point. Future work will include the option to simulate and interpolate input data to enhance the accuracy of the LSTM and explore suitable approaches to mitigate, tolerate or exploit uncertainty through deep learning.
Highlights
Spatial geometry enables uncertainty to be forecast where data is scarce, but is not able to extrapolate such forecasts with the introduction of new data over time. This is a feature of Long short-term memory (LSTM), which, if combined with spatial geometry, can enable multistep prediction of uncertainties given by sporadic data
For the size of the sample dataset used in this study and the development stage of the conceptual model, the LSTM network was not able to make predictions of cost variance using the end coordinate variables or symmetry in the state space that would enable confident and accurate forecasting
Incorporating deep learning LSTM networks with spatial geometry, the dynamic multistep prediction model analyses the geometric symmetry of given input variables through polar force-fields in vector space
Summary
1u.tIilnMistrionodgdeurcncotnipodrniotiaocntivmeonmitaoirnitnegnaanncde dsetreaptelgeiaersninagre tograacdhuiaelvlye cdctsdcwmeuutteruoacrutteuoachoohiehohoeehnagqnoqnosnntsppoammmeeiccDdiriiiddururccttlrlcnnstksoorriilipameteeiiaffmmueeesheeeggepprrnnaarrelccinccddentuttimmccnuudtcuuoosaazttgeeieddniiigyiinn,mmtssiiaseeddlledvvaaiionniictcttoncntseetttiirttifttceessopppaaonnaaaettionyypporroonttrrrtwggscaantteeiieeooondrruaattffnu..wvvo,addhhrrrrdduiiltttsaanooaarntnhiiirriimHHuicddtsiidmmtccooddtmmtottellhiceeioiittoosaaauuiinomaeeiieetppiccdnbbwwoogggnnssnrc--laaeehhiauiennbbhhidofeezllllnnccfiiydfassiillaaoodmssvvmateaaddattyy’shussttuuyiiytcieellscaaaooteenniopmeclltt[[..yrrotbbnrrddnt11e,,uffttAcAiittniillmoouutic–c–hhorneedlbbdi,tmmeirrrro33teeonieffaaaennnossbaaoon]]ccvraassrvg.hmmaassro..inengrrkooiinneesnyyrreIsseeifTTddoennlaayyeoolnn,,gssdccootaiiiphhttnttuugwwaannniipeeapnnliirtnnssnnssallttgssmemmdohhmeeeyyaotteeuuddnrdndwweeopptssnntiiooepdettrrssvittuvtllrraanuuiitrnteeeeriioosicmmheiananiffetddssdmmrtoiseetmmmmrecconzaaecceeenaassdeefedntteaaccoos,,aann,sraahseipmllyyeeiittoaddntaaccvnnaeeeqqhcccmcmorrttuuedrssettrlrlleuuooooeeeefeneellewliiuervviiasasoggnn..attppeettottmfaahsffomompaadrtiiOOiimmaoo:rrirnnneioonnrrmimatcsyylaarrgooinnggcconneedceeiniiieaiuuaeeccnennnnn.ccrtgoonnsseediogttattgaaddaazTrrtiaaocaaussffniananqqtiiddeeitoohttnovvmognntttddimuusseeoaarriicaanmameettsaaaoaeaeggriiptiiaddiinnllccineeimmegrrsseecoaamntnaaefefttasnrlnnhbbiiorouuauaoossoeeesffittssaiaiittttttffiiiscsinnnabelluuiiiiuuttnnnnneeoiirnnnoodeevaoorrrtrtdddggdgofggtnnnyynl,eeaaeeeff,, fouuttucroemeeve(nBtawyietshiaonr),wwithhoiluet aprpiorredinicfotiromnaitsioann[4in,5d]i.cation of a pbpmoamnHSlofSloSSptaSSptaseeehhrurfurroffrroaeeeeeesooaaweeteeteeOictecccccwduddrrAAtaathddnattttttmnhnhfferuu..npiiiiiieleiirrnneeieoiooooolscccccvaabrnnlonnnnnnhtttoittpemmelggiihppnseeaefdoovrrt,e22rrgxxaeeo44rod,33nneuaaooaa,,maawwad..nnn..cpbppicammswwccunttddloorSSieootTTnsiooloecwfrriiiimeeeissdthhaffvvttnnartkkdeheeccmhhooiibeeeemaaisuddsitttratrraarrlsttiihbtaecaeeppeiiooniiilliffnooldtlnnmmorsocccrrorrannyiteygnnaaggeeaasnnefolrpoo:ayssslleomm55srggiiwrddeetttaem.s(roommeiioeexxdadinieeigwwnnIiff)mesdiwwnllttiiiiiiargglnhssnnssiiuotdylct.tttooissttcchytoaahhhhcuscciiaettIuurrerrznnulrrteecoopnnkdkeyyssuucceggnthnnernssvoocctbpoaarrpdacceeuhiirreaattnneernessrwnreeeessuumilnnoissaccigs,ssrrdnrrooeeddttcdllananuubeeehhpuucnaabrtftuppiilleeeeeeoossrrrunnottsoicppaaroccrssiimmnennggtrtoonnssicllhhdorfstiicttfddonneetteoevoerruuutfaootrirssmeexxddinrbbnncmcggcoccnnmitdltaaccaasiipoiis,aanysggillttccnneenntaallllltoomtttimkkccirrouuhgghdddtrhhntettiuuggesspoafeiaaossagsssttfffttllrrtoiiceeaannrrpfuuaamooaaanncaapeclnccmmuttrttriuunntto[ttiirluuraohheeeyyfeeooa4ddnnppopianrrnndn..msnnnh,ddeelldectiit5uslltteeAAssasqqstiiuhhaicc]wwmmurcwuuiiic.eeonnahtiidddssnraseooesstnriiaeilaallttoossmmtrrytdflydfaattmmhggkkoleavveatteiituuiix(miiaeessnniaaoavvdniioosillneeirrtnnddintteeiegsiunnvvf)ttddiineeldtnnssifssterssiiitteeintwgitt/eeeehhnhneeeeoidooliiiiiiswwsiiogneeeppnnnnppnnrss..,ff type of components (e.g. mechanical, electrical, electronical). Classical methodologies considering mainly single products product varieties (high-volume to low-volume production) [1]. Solitary, already existing product families analyze the T2To2h1ics2o-is8p2ae7n1ow©pietn2h0a2ct0hceiTsshseaaurAtigucmtlheoeurnnsd,tiPenrugtbhleivsChaerCdieBbtyYyE-NalssCev-wNieeDrlBlli.cVaes.nsteo(hbtetp:a//bcrleeattiovecommproonds.uorcgt/lsictreunscetsu/brey-nocn-nad/p4.h0y/)sical level (components level) which idP22eee1nr2tr-i8efv2yi7e1wp©uon2sd0se2irb0tlhTeehereoAsppuothinmsoirbiszi,laPitutyiboolfinsthheedpsobciyteenEntlitsfieiacvlicseormBim.nVit.tetehoef theexCisIRtiPneg2020 cGaloubsaelsWedbiCffoincfuelretinecse regarding an efficient definition and pTrhoidsuiscatnioonpesnyasctceemss,airtticisleiumndpeorrtthaenCtCtoBYha-NvCe-aNDprleicceinssee (khnttop:w//clreedagtiveecommcoomnsp.oargr/ilsicoennseos/fbyd-nicf-fnedr/e4n.0t/) product families.
Sign-in/Register to view highlights & summary 
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call