Terminology as the Basis for Building Engineering Feature-based Models

186 Rica do Ei o B un Te minology as he Basis o Building Enginee ing Fea u e-based Models doi.o g/10.35321/ e m27-08 Te minology as he Basis o Building Enginee ing Fea u e-based Models RicaRdo Ei o BRun Uni e sidad Ca los III de Mad id ABSTRACT Sa elli e ope a ions equi e he combined use o di e en ools o suppo en- ginee ing ac i i ies and o con ol he spacec a . This communica ion is man- aged by he Moni o ing and Con ol Sys em (MCS) ha ecei es eleme y da a om he spacec a and eleases elecommands o keep he sa elli e’s a i ude and ligh pa h. These complex sys ems a e de eloped as open pla o ms ha can be ex ended and cus omised o suppo mission-speci ic equi emen s and objec i es. As a gene al ule, i can be s a ed ha hese so wa e applica ions a e good candida es o implemen ing a iabili y mechanisms in a s uc u ed, planned way and ha hei unc ionali y is a good candida e o analyse he ea- sibili y o applying ea u e-based modelling echniques. This pape desc ibes he use o e minology analysis o build a ea u e model o suppo equi e- men s analysis o his ype o so wa e-based sys ems. KEYWORDS: Technical e minology, Fea u e-based models, Ae ospace enginee ing, Te mi- nology ex ac ion ANOTACIJA Palydo ų ope acijoms eikalingi į ai ūs į ankiai, ski i už ik in i sklandų inži- ne inį da bą i aldy i e d ėlai į. Šį yšį aldo S ebėjimo i kon olės sis ema (SKS), ku i gauna eleme inius duomenis iš e d ėlai io i duoda elekoman- das, kad palaiky ų palydo o padė į i sk iejimo ajek o iją. Šios sudė ingos sis- emos y a suku os kaip a i os pla o mos, ku ias galima išplės i i p i aiky i pagal konk ečios misijos eikala imus i ikslus. Galime eig i, kad pap as ai šios aikomosios p og amos gali bū i naudojamos no in s uk ū uo ai i pla- nuo ai įdieg i a ian iškumo mechanizmus, o jų unkcionalumas leidžia anali- zuo i požymių modelia imo me odikų p i aikymo galimybes. Šiame s aipsnyje ap ašomas e minologinės analizės panaudojimas požymių modeliui, palaikan- čiam eikala imų analizę šio ipo p og aminėms sis emoms, suku i. ESMINIAI ŽODŽIAI: echninė e minologija, požymių modeliai, a iacijos inžine ija, e minų a pažinimas 187Te minologija | 2020 | 27 1. INTROduCTION NASA glossa y de ines sa elli es as: “a ee- lying objec ha o bi s he Ea h, ano he plane , o he sun.1” Sa elli es a e a ype o spacec a ha a els in a egula , clea ly de ined o bi a ound he cen e o g a i y o ano he celes ial body (Ga ne 1996: 4). Since he launch o he i s a i icial sa elli es – Spu nik 1 on Oc obe 4 h, 1957 and Explo e 1 on Janua y 31s , 1958, a la ge numbe o sa elli e missions ha e been launched wi h di e en pu poses: as onomical explo a ion, p o ision o communi- ca ion and na iga ion se ices, ea h obse a ion, econnaissance, and scien i ic missions. Sa elli es a e complex ae ospace sys ems made up o g ound- and space-based elemen s: he spacec a mus be ope a ed by a con olling elemen on Ea h and emain in con ac wi h i . Today, sci- en i ic p og ess and se ices o use s depend on sa elli es and sa elli e cons ella ions. Rele an examples include he Hubble Space Telescope (HST), na iga ion sys ems like GPS (Global Posi ioning Sys em), GLONASS and he Eu opean Galileo, UARS (Uppe A mosphe e Resea ch Sa elli e) o GOES (Geos a iona y Ope a ional En i onmen Sa elli e). The o al num- be o sa elli es launched since 1957 – acco ding o he US SSN2 ca a- logue – Is close o 18200; uNOOSA’s 2017 Index o Objec s Launched in o Ou e Space epo s 4635 sa elli es cu en ly o bi ing he plane , wi h an inc emen o 357 sa elli es (8.95%) conce ning he p e ious yea 3. The pu pose o his a icle is o demons a e he need o applying e minol- ogy managemen and ex ac ion o suppo he de elopmen o ea u e- based models o o ganize he concep s ha desc ibe he unc ions o he so wa e applica ions used o sa elli e moni o ing and con ol. Sa elli es a e classi ied by pu pose and ype o o bi . The pu pose e e s o he se ices he sa elli e is in ended o p o ide. Rega ding he o bi , a dis inc ion is made be ween Low Ea h O bi (LEO), Medium Ea h O bi (MEO), and Geos a iona y O bi (GEO). LEO sa elli es – used o science and Ea h obse a ion – ollow an ellip ical o bi ; as hei isibili y om g ound s a ions is limi ed, da a a e s o ed on-boa d and sen o he g ound 1 See h ps://www.g c.nasa.go /www/k-12/TRC/lae s/lae s_s.h ml#sa elli e [accessed 2020-08-01]. 2 The US Space Su eillance Ne wo k is in cha ge o he de ec ion, acking, ca aloguing and iden i ica ion o a i icial objec s o bi ing he ea h. The ca alogue is a ailable a : h ps://www.space- ack.o g/#/ss [accessed 2020-08-01]. 3 Uni ed Na ions O ice o Ou e Space A ai s. See h ps://www.pixaly ics.com/sa s-o bi ing-ea h-2017/ 188 Rica do Ei o B un Te minology as he Basis o Building Enginee ing Fea u e-based Models s a ions when he ai c a becomes isible. GEO sa elli es a e used o elecommunica ion and me eo ological pu poses; elecommunica ion sa - elli es ecei e adio equency (RF) signals om he Ea h, ampli y hem, shi hei equency, and ansmi hem back o he Ea h. They o bi 36000 km abo e he Equa o (in he same plane), and hei o a ion is synch onous wi h he o a ion o he Ea h, s aying abo e he same poin a he Equa o all he ime)4. Simila o sa elli es, deep space scien i ic missions also need simila moni o ing and con ol o elecommand, and eleme y ecep ion. Sa elli e missions equi e dedica ed s a o comple e eal- ime, con- inuous moni o ing o he s a us and posi ion o he sa elli e. Mission con ol is he se o asks execu ed a e launch by ope a ion enginee s wi h he help o so wa e applica ions, and in ol es he exchange o da a be ween he g ound and space segmen s o moni o ing and con ol he s a us o he sa elli e’s onboa d subsys ems. In scien i ic missions, i is also necessa y o ecei e he in o ma ion om he sa elli e payload and deli e i o he end-use s (scien i ic communi y). Typical unc ions ex- ecu ed du ing mission con ol include he ecep ion and analysis o e- leme y, elecommanding, and acking. Teleme y is he da a ansmi ed om he sa elli e o he Ea h in o ming o he s a us and condi ions o he sa elli e and i s subsys ems. Telecommands a e he o de s ansmi ed om he g ound o he spacec a o con igu e and ope a e i . (uhlig, Sellmaie , and Schmidhube 2015: 232). T acking and s a ion keeping, also known as anging, consis s o he moni o ing and de e mina ion o he ligh pa h and posi ion using RF echniques and signals. F om a so wa e enginee ing pe spec i e, sa elli e con ol equi es di - e en applica ions and ools o ligh dynamics, mission planning, elem- e y and elecommanding, ne wo k con ol and ou ing, as well as in e - aces be ween hem. This complexi y has led o he p o ision o di e en solu ions by space agencies. One o he mos ele an miles ones in he de elopmen o MCS so wa e was he decision o he ESA G ound Sys ems Enginee ing depa men o de elop and license o he Eu opean indus y a se o so wa e applica ions dis ibu ed unde he name MICONYS® 4 GEO sa elli es a e dis ibu ed in a limi ed a ea in space. O bi slo s a e assigned by he In e na ional Telecommunica ion Union (ITu) o a oid collisions, a isk ela ed o he space deb is opic ha is ecei - ing g ea e a en ion oday. 189Te minologija | 2020 | 27 (Mission Con ol Sys em). MICONYS and i s SCOS-2000® componen a e p obably he bes -known examples o his ESA’s policy o so wa e de elopmen and inno a ion and echnology ans e (Kau ele , Jones and Ka l 2001). SCOS-2000 suppo s elecommanding, eleme y ecep ion, display, and a chi al. SCOS-2000 was he esul o he expe ience acqui ed by ESA in he de elopmen and ope a ion o p e ious simila sys ems: MSSS, SCOS-1, and SCOS-2. Today, ESA is de eloping he new MCS sys em, aimed o eplace SCOS-2000 in he u u e. I s name is Eu opean G ound Sys ems – Common Co e (EGS-CC). Simila o SCOS, he p ojec has he pu pose o de eloping a common in as uc u e o suppo he moni o ing and con ol o space missions in he p e- and pos -launch phases, using mode n echnologies and se ice-o ien ed a chi ec u es (Pec- chioli e al. 2012). The de elopmen o EGS-CC is no only unde ESA esponsibili y: Eu opean na ional space agencies (CNES, UK Space Agen- cy, and DLR) and indus ial companies (AIRBUS De ence and Space, Thales Alenia Space and OHB Sys ems) a e pa o he p ojec . Besides ESA p ojec s, he e a e o he ini ia i es aimed o de elop a gene ic MCS so wa e sys em. NASA has comple ed simila p ojec s, mos o hem in-house de elopmen s. The Godda d Space Fligh Cen e de el- oped wo sys ems, ITOS and ASIST, o managing missions like WMAP (Wilkinson Mic owa e Aniso opy P obe), IMAGE (Image o Magne opause- o-Au o a Global Explo a ion), EO-1 (Ea h Obse ing 1), ST-5 (Space Technology 5), SdO (Sola Dynamics Obse a o y) o LRO (Luna Recon- naissance O bi e ) (P a e al. 2007). In all hese cases, we a e dealing wi h a complex sys em ha needs o implemen di e en unc ions ha can be ac i a ed o no depending on he sa elli e and he mission’s cha ac e is ics. due o ha , hese so wa e- in ensi e sys ems a e good candida es o be de eloped using p oduc -lines and ea u e-based modelling echniques, which ely on a clea and well- o ganised o ganiza ion o concep s o unde s and he domain and sys em needed capabili ies. 2. FEATURE-bASED MODELLING AS AN ORGANIzATION OF CONCEPTS Fea u e-based modelling is one o he echniques applied in so wa e p oduc -line enginee ing, a discipline o building eusable so wa e p o- g ams ha became popula a he end o he nine ies. SPLE in ends o 190 Rica do Ei o B un Te minology as he Basis o Building Enginee ing Fea u e-based Models build eusable componen s and a e ac s ha can be la e combined o build p oduc s sui ed o he needs o a speci ic clien and con ex . Expe- iences in SPLE a e widely documen ed in he p o essional and aca- demic li e a u e. Capilla (2013) included se e al case s udies whe e SPLE was applied wi h success: Boeing’s ope a ional, mission-c i ical ligh p o- g ams o a ionics and cockpi unc ions, Bosch’s engine-con ol so wa e o gasoline sys ems, Hewle Packa d’s p in e so wa e, Toshiba’s powe gene a ion, and ansmission equipmen and Gene al Mo o s’ con ol so - wa e o powe ains. The bene i s o so wa e p oduc lines (SPL), ac- co ding o Apel e al. (2013: 9), include he ailo ing o so wa e p oduc s o he speci ic needs o he clien s, educed cos – as a se o eusable asse s can be combined in di e en ways o gene a e new p oduc s -, imp o ed quali y and ime- o-ma ke . Fea u es and ea u e-based modelling a e ele an concep s in he de- elopmen o SPL. ISO/IEC/IEEE 24765:2010, Sys ems and so wa e enginee ing — Vocabula y, o e s a gene al de ini ion o ea u es, aken om IEEE S d 829:2008 IEEE S anda d o So wa e and Sys em Tes Docu- men a ion: “A dis inguishing cha ac e is ic o a sys em i em. NOTE: includes bo h unc ional and non- unc ional a ibu es such as pe o mance and eus- abili y.” Kang and Lee (2013, 28) de ine ea u es as “abs ac concep s e ec i ely suppo ing communica ion among di e se s akeholde s o a p oduc line, and he e o e, i is na u al and in ui i e o people o exp ess commonal- i y and a iabili y o p oduc lines in e ms o ea u es.” Fea u es se e di e en pu poses in he p oduc idea ion and de elop- men p ocess. They a e means o communica e he p oduc cha ac e is ics and suppo he iden i ica ion o equi emen s; hey a e also he concep s ha guide design and implemen a ion decisions. The de elopmen o a p oduc -line comp ises wo complemen a y li e cycles: • domain enginee ing, and • Applica ion enginee ing. IEEE 1517-2010 s anda d de ines domain enginee ing as: “Li e cycle consis ing o a se o p ocesses o speci ying and managing he commonal- i y and a iabili y o a p oduc line. domain enginee ing analyses he domain o a p oduc line and de elops a se o eusable a e ac s. These a e ac s include so wa e equi emen s, design elemen s, es cases and 191Te minologija | 2020 | 27 p ocedu es, use documen a ion, e c. domain analysis can be seen as a so o equi emen s enginee ing o he whole p oduc line, including he iden i ica ion o an icipa ed a iabili y. The main a e ac gene a ed by domain analysis is he ea u e model ha will speci y and desc ibe he p oduc s wi hin he line. Fea u e modelling is a diag amming echnique ha was in oduced in he nine ies wi h he FOdA (Fea u e-O ien ed Domain Analysis) me hod- ology. FOdA p o ided p imi i es o ep esen ing s uc u al ela ions (composi ion, gene aliza ion, and specializa ion), op ionali y, al e na i e- ness, and mu ual dependencies. I was la e e iewed by di e en au ho s (Kang and Lee 2013: 30–31). Fea u e diag ams a e he isual ep esen a- ion o ea u e models, whe e ea u es a e ep esen ed as boxes in a hi- e a chical ee. Each node has an a ached label wi h he name o he ea u e. The hie a chical a angemen o he ea u es c ea es pa en -child ela ionships. I a child ea u e is selec ed, i s pa en mus also be se- lec ed. diag ams can make a dis inc ion be ween manda o y and op- ional ea u es, and iden i y he combina ions o ea u es ha a e alid. In pa icula , ea u e diag ams can ep esen : • Abs ac ea u es, which a e used o o ganise he ea u es in he ee bu a e no bound o implemen a ion a e ac s. They a e ep- esen ed wi h g ey boxes. • Conc e e ea u es, which co espond o implemen a ion a e ac s and a e ep esen ed wi h whi e boxes. • Manda o y ea u es, which ha e a illed bulle on op o he uppe bo de o hei box. • Op ional ea u es, which ha e a non- illed bulle on op o he up- pe bo de o hei box. • The need o selec ing jus one o he child ea u es o a speci ic pa en (exclusi e OR, XOR, o one-ou -o -many). I is ep esen ed wi h an emp y a c a he lowe bo de o he pa en ea u e’s box. • The possibili y o selec ing mo e han one child ea u es o a spe- ci ic pa en (OR o some-ou -o -many). I is ep esen ed wi h a illed a c a he lowe bo de o he pa en ea u e’s box. • dependencies be ween ea u es, ep esen ed by a ows wi h ex ual anno a ions. 192 Rica do Ei o B un Te minology as he Basis o Building Enginee ing Fea u e-based Models The diag am below shows a ypical ea u e diag am wi h he con en- ions desc ibed abo e: Fig. 1. Fea u e diag amming example (Sou ce: Ga gan ini 2015) A ea u e model should include in o ma ion addi ional o he diag am. Apel e al. (2013: 27) indica e he possibili y o adding hese da a: • “Desc ip ion o a ea u e and i s co esponding se o equi emen s. • Rela ionship o o he ea u es, especially hie a chy, o de , and g ouping. • Ex e nal dependencies, such as equi ed ha dwa e esou ces. • In e es ed s akeholde s. • Es ima ed o measu ed cos o ealizing a ea u e. • E c.” The de elopmen o a ea u e model o ep esen he unc ional cha - ac e is ics o so wa e applica ions o sa elli e con ol mus s a om a clea unde s anding and modelling o he concep s ha make up he domain. To his end, ac i i ies ela ed o e minology and e minog aphy, he iden i ica ion o e ms, concep s, and hei ela ionships p o ide he basis o build he a ge model. 193Te minologija | 2020 | 27 3. WORK METHODOLOGy The p oposed ea u e-model has been comple ed ollowing hese s eps: • Iden i ica ion and e iew o p o essional and academic li e a u e published in his a ea. This in ol es sea ching o in o ma ion abou he app oach ollowed in ae ospace p ojec s led by en i ies like ESA (Eu opean Space Agency) o NASA (Na ional Ae onau ics and Space Adminis a ion). • de elopmen o a glossa y based on he analysed li e a u e, apply- ing e minology managemen echniques o eco d in o ma ion abou e ms, ela ionships be ween hem, de ini ions iden i ied in he documen s, and he con ex whe e he e ms a e used. • C ea ion o a ea u e-based model ha ep esen s he unc ions o - e ed by sa elli e moni o ing and con ol so wa e applica ions, us- ing he glossa y as a basis. The inpu s used o iden i y e ms included a subse o echnical docu- men s ha desc ibe he selec ed p oduc : ope a ion manuals, whi e pape s, and aining ma e ials. The ea u e model is a hie a chical ee o ea u es ma ked as manda o y o op ional, wi h dependencies be ween hem. In he case o he so wa e applica ion unde analysis, besides he iden i ica- ion o op ional and manda o y ea u es, addi ional p oduc line a iabil- i y equi emen s we e iden i ied. In pa icula , some o he unc ions suppo ed by he so wa e unde analysis equi e o e w i ing o cus om- izing exis ing code. The iden i ica ion o hese a iabili y cases was made wi h he suppo o expe s who de elop hei ac i i y in he de elopmen and cus omiza ion o his ype o so wa e applica ion. Pe sonal in e iews we e made o collec ha in o ma ion. The ool selec ed o c ea e he ea u e model and diag ams is Fea u eIDE5. This is an open-sou ce ool based on Ja a and Eclipse, de eloped by s a a he O o- on-Gue icke-Uni e si ä Magdebu g. Wi h Fea u eIDE, i is possible o c ea e a ea u e model using a g aphical edi o , ma k ea u es as manda o y, op ional, o abs ac , and build he hie a chical ee. Once he ea u e model is buil , you can c ea e di e en con igu a ions: selec- ions o ea u es ha will be used o gene a e he a ge so wa e applica- 5 See h ps://Fea u eIdE.gi hub.io/. Las checked: 01-03-2018. 194 Rica do Ei o B un Te minology as he Basis o Building Enginee ing Fea u e-based Models ion sou ce code. un o una ely, he ool does no p o ide capabili ies o manage e ms o e minology uni s, which makes necessa y he use o complemen a y ools. Fig. 2. Fea u eIDE (c ea ed by he au ho ) 4. PRESENTATION OF RESULTS AND DISCUSSIONS This sec ion summa izes one sec ion o he ea u e model o Mission Con ol Sys ems (MCS) so wa e applica ions. The analysis o he concep s ex ac ed om he documen s led o an o ganiza ion o he sys em unc- ionali ies in o hese a eas: • Desk op and Session Managemen , which p o ides he unc- ions o log in, s a a session, and launch he di e en applica- ions. In he case o managing mul iple sa elli es, he use will be able o swi ching be ween sa elli es’ wo kspaces. • Teleme y chain, which includes Teleme y p ocessing, Ala m Manage , and Teleme y display.