Prikaz osnovnih podataka o disertaciji

Development and implementation of multimodal system for attention monitoring in naturalistic work environments.

dc.contributor.advisorMačužić, Ivan
dc.contributor.otherFilipović, Nenad
dc.contributor.otherTodorović, Petar
dc.contributor.otherKović, Vanja
dc.contributor.otherMilovanović, Miloš
dc.creatorMijović, Pavle
dc.date.accessioned2017-03-06T09:37:52Z
dc.date.available2017-03-06T09:37:52Z
dc.date.available2020-07-03T15:12:34Z
dc.date.issued2016-11-14
dc.identifier.urihttps://nardus.mpn.gov.rs/handle/123456789/7673
dc.identifier.urihttp://eteze.kg.ac.rs/application/showtheses?thesesId=4636
dc.identifier.urihttps://fedorakg.kg.ac.rs/fedora/get/o:762/bdef:Content/download
dc.description.abstractKako tehnologija stalno napreduje, industrijske nesreće koje se vezuju za neispravnost tehničkih sistema su skoro skroz umanjene. Iz tog razloga, ljudska greška se smatra uzročnikom oko 80% nesreća u industriji. Jedan od glavnih uzročnika ljudske greške je limitirana mentalna izdržljivost ljudskih operatera, koja uzrokuje pad u pažnji radnika i posledično vodi do grešaka u radu. Klasične ergonomske metode koje se koriste za procenu kognitivnog stanja čoveka su uglanom kvalitativne i subjektivne, i prema tome su prilično nepouzdane. Iz tog razloga, psihofiziološki senzori su počeli da se primenjuju u ergonomiskim istraživanjima, sa ciljm da obezbede objektivne i kvantitativne mere radnikovog kognitivnog stanja. Prateći taj trend, neuroergonomija se pojavila kao naučna pod-disciplina ergonomije. Prednost korišćenja neuroergonomskih metoda, je u tome što neuroergonomija istražuje funkcionalnu zavisnost između dinamike mozga i bihevioralnih parametara i tako zaobilazi teoretske veze koje opisuju korelaciju između ovih parametara, a koje su korišćene u ergonomiji. Ova disertacija predstavlja naučni okvir za multimodalni sistem koji je predložen da se koristi za praće pažnje radnika i koji koristi psihofiziološke senzore i bihevijoralna merenja. Sistem se sastoji od psihofizioloških senzora, kao što su: galvanski reakciju kože (eng. galvanic skin response - GSR), merenje otkucaja srca (eng. heart rate –HR) i elekroencefalografiju (eng. Electroencephalography – EEG); bihevioralne modalitete kao što su: Vremena reakcija (eng. reaction times – RTs) i senzore za praćenje pokreta (eng. motion capture – MoCap), “Kinect” the “Leap Motion”. Iako je predstavljen okvir za snimanje pomenutih modaliteta u realnom vremenu, ova disertacija je fokusirana na rezultate koji su dobijeni snimanjem EEG, RTs i Kinect modaliteta. Glavni cilj disertacije je istraživanje mogućnosti korišćenja savremenog prenosnog EEG-a u industrijskim uslovima, sa ciljem praćenja pažnje radnika. Prethodna istraživanja koja su koristila EEG su bila uglavnom obavljana u kontrolisanim laboratorijskim uslovima i zbog toga, nalazi iz tih studija se uzimaju sa određenom dozom rezerve. Da bi se snimio EEG u realnom radnom okruženju, radno mesto u kojem radnici sklapaju hidraulično crevo je verodostojno replicirano i subjekti u studiji su simulirali taj proces. Disertacija se sastoji od četiri eksperimentalne studije. U prvoj studiji, ispitivano je kako česte mikro-pauze utiču na nivo pažnje radnika, poredeći amplitude P300 Komponente evociranih kognitivnih potencijala (eng. event-related potential – ERP) pre i neposredno posle perioda mikro-pauze. Glavni nalaz je da mikro-pauze pozitivno utiču na nivo pažnje radnika i predloženo je njihovo uključenje u dnevne aktivnosti radnika. U drugoj studiji, istraživano je da li radnici imaju veći nivo pažnje ukoliko im je nametnuto sa kojom rukom da počnu sklapanje creva. Dve psihološke paradigme su bile predstavljene učesnicima u studiji, paralelno sa simuliranm akcijom sklapanja creva. U prvoj paradigmi, učesnici su mogli da izaberu da otpočnu operaciju sa bilo kojom rukom, dok su u drugoj bili uslovljeni da započnu operaciju rukom koja odgovara smeru strelice koja se prikazivala na ekranu ispred njih. Ovo istraživanje je otkrilo da su učesnici imali veći nivo pažnje u slučaju uslovljavanja kojom rukom da započnu operaciju, jer je amplituda P300 komponente bila značajno viša u poređenju sa slučajem kada su mogli slobodno da izaberu sa kojom rukom će započinjati zadatak.sr
dc.description.abstractAs technology is ever advancing, industrial accidents related to technological malfunctioning have been almost diminished, leaving the human error responsible for up to 80% of the remaining accidents. One of the main causes for this is limited mental endurance of human operators’, which causes the attention decline and consequently leads to an operating error. Classical ergonomics methods for assessing the operators’ cognitive state are still dependent on the subjective and qualitative methods, thus being unreliable. For that reason, in the recent years the psychophysiological sensors were included in the ergonomics research, with the aim of providing the objective and quantitative measures of the operators’ cognitive state. Following that path, the neuroergonomics emerged as a scientific discipline, which investigates the human brain functions in relation to performance at work. The advantage of using neuroergonomics is that it investigates the functional relationship between brain dynamics and behavioral parameters, thus avoiding theoretical constructs that describe the correlation between these two, and which are ubiquitously used in ergonomics research. The present dissertation introduces a framework for the multimodal attention monitoring system, utilizing psychophysiological and behavioral measurements. The multimodal system consists of psychophysiological sensors, such as galvanic skin response (GSR), heart rate (HR) sensor and electroencephalography (EEG), the behavioral modality of the reaction times (RTs), and the motion capture (MoCap) sensors Kinect and the Leap Motion. Although the framework for synchronous and real-time recording for all the sensors was provided, this thesis was focused solely on the results obtained from the EEG, RTs and Kinect recordings. The main aim of the presented dissertation is to investigate the possibility of utilization of the recently available wearable electroencephalography (EEG) in industrial setting, with the goal of the operator’s attention monitoring. Previously reported EEG studies that were concerned with the attention states of the operators were mainly confined to the strictly controlled laboratory conditions and therefore, the findings from these studies needed to be taken with the certain ambiguity. In order to record the EEG in naturalistic environment, specific workplace where operators’ assembly the hoses, used in hydraulic break systems in vehicles, was faithfully replicated and the participants in the studies simulated the manual assembly operations. The present dissertation consists of four experimental studies, where the first two were concerned with investigation how different work conditions influence the cognitive state of the operators’, i.e. the studies were concerned with the assembly task design. In the first study, the influence of the frequent micro-breaks on the cognitive state of the workers’ was investigated, by comparing the P300 event-related potential (ERP) amplitude prior and immediately following the micro-break period. It was found that the micro-breaks enhance the attention of the operators’ and the proposal for their inclusion in the regular work routine was made. Second study investigated the influence of hand alteration on the attention level of the operators’. For that aim, the participants in the study were presented with two distinct task: the one in which they could initiate the assembly operation with whichever hand they preferred, and the one in which they were conditioned with which hand they should initiate the operation. This study revealed that the instructed hand responding induces the higher attention, as assessed through the P300 component’s amplitude, compared to the experimental condition where the participants could freely choose the hand for the initiation of the assembly operation. Further, a framework for the on-line assessment of the operators’ cognitive state was provided. In the third experimental study, the propagation of the P300 component’s amplitude was observed and correlated with the RTs. On the group level, a negative correlation was found, confirming the previously reported finding. However, due to individual differences, the correlation on the individual level was inconsistent, emphasizing the necessity for the individualized EEG measurements for the reliable attention monitoring system. Finally, it was investigatetd whether the quantity of task unrelated movements corresponds to attention of the operator, as previously shown to be negatively related to the attention of operators’. For that aim, the concept of movement energy (ME) was introduced and correlated with EEG attention-related modalities. The initial finding from this study showed that the ME is negatively related to the EEG attention-related modalities and proved that the future attention monitoring system can be built based on these modalities.en
dc.formatapplication/pdf
dc.languagesr
dc.publisherУниверзитет у Крагујевцу, Факултет инжењерских наукаsr
dc.relationinfo:eu-repo/grantAgreement/EC/FP7/289837/EU//
dc.rightsopenAccessen
dc.rights.urihttps://creativecommons.org/licenses/by-nc-nd/4.0/
dc.sourceУниверзитет у Крагујевцуsr
dc.titleRazvoj i implementacija multimodalnog sistema za praćenje pažnje radnika u realnom radnom okruženjusr
dc.title.alternativeDevelopment and implementation of multimodal system for attention monitoring in naturalistic work environments.en
dc.typedoctoralThesisen
dc.rights.licenseBY-NC-ND
dcterms.abstractМачужић, Иван; Ковић, Вања; Миловановић, Милош; Филиповић, Ненад; Тодоровић, Петар; Мијовић, Павле; Развој и имплементација мултимодалног система за праћење пажње радника у реалном радном окружењу; Развој и имплементација мултимодалног система за праћење пажње радника у реалном радном окружењу;
dc.identifier.fulltexthttp://nardus.mpn.gov.rs/bitstream/id/48121/Disertacija.pdf
dc.identifier.fulltexthttps://nardus.mpn.gov.rs/bitstream/id/48121/Disertacija.pdf
dc.identifier.fulltexthttps://nardus.mpn.gov.rs/bitstream/id/48122/bitstream_48122.pdf
dc.identifier.rcubhttps://hdl.handle.net/21.15107/rcub_nardus_7673


Dokumenti za doktorsku disertaciju

Thumbnail
Thumbnail

Ova disertacija se pojavljuje u sledećim kolekcijama

Prikaz osnovnih podataka o disertaciji