《人体工效学 (2).pdf》由会员分享，可在线阅读，更多相关《人体工效学 (2).pdf（10页珍藏版）》请在taowenge.com淘文阁网|工程机械CAD图纸|机械工程制图|CAD装配图下载|SolidWorks_CaTia_CAD_UG_PROE_设计图分享下载上搜索。
1、International Journal of Industrial Ergonomics 33(2004)181190Measurement consistency and three-dimensionalelectromechanical anthropometryDavid J.Feathers*,Victor L.Paquet,Colin G.DruryDepartment of Industrial Engineering,School of Engineering and Applied Sciences,State University of New York,342 Bel
2、l Hall,Buffalo,NY 14260-2060,USAReceived 28 March 2003;received in revised form 1 August 2003;accepted 20 August 2003AbstractTwo pilot studies investigated potential sources of error in static human body surface measurement for conventionalanthropometry methods and three-dimensional electromechanica
3、l methods under different experimental conditions.Inthe first pilot study,two anthropometrists measured Humeral breadth,Radiale-Stylion length and Wrist breadth of acadaveric arm repeatedly in nearly nude and lightly clothed conditions with conventional and electromechanicalapproaches in two separat
4、e,repeated measurement sessions.Analysis of variance(ANOVA)performed on thesemeasurements demonstrated significant differences across measurers,methods,the interaction between measurer andclothing,and the interaction between measurer and methods,suggesting systematic error contributions for thesevar
5、iables.ANOVA performed on the standard deviation of data for each anthropometric dimension showed differencesacross methods and clothing conditions,demonstrating differences in measurement consistency for these variables.Inthe second pilot study,measurement consistency was evaluated for the conventi
6、onal and electromechanical methods foranthropometric measurements of ten wheelchair users who were clothed and not capable of maintaining erect seatedpostures for the measurement session.The measurement consistencies for repeated measurements of acromion height,biacromial breadth,eye height,knee hei
7、ght and waist depth obtained with each method were compared to establishedstandards.ANOVA showed differences between methods and measurers for some of the anthropometric dimensions,although the magnitude of the differences was relatively small.Relatively low variability in measurements withinmethod
8、for each dimension within condition was found in both studies.This suggests that conventional dimensionsrecorded with three-dimensional electromechanical approaches can be measured consistently,at least for theanthropometric dimensions and experimental conditions considered in these studies.Relevanc
9、e to industryAnthropometric data provides a valuable source of information to ergonomists and designers who attempt toconsider a range of body sizes and abilities in the design of occupational environments and products.These pilot studiesinvestigate the reliability of anthropometric data collected w
10、ith conventional and new three-dimensional measurementapproaches.r 2003 Elsevier B.V.All rights reserved.Keywords:Anthropometry;Electromechanical measurements;Reliability;Wheelchair usersARTICLE IN PRESS*Corresponding author.E-mail address:featherseng.buffalo.edu(D.J.Feathers).0169-8141/$-see front
11、matter r 2003 Elsevier B.V.All rights reserved.doi:10.1016/j.ergon.2003.08.0011.IntroductionThere is no correct anthropometric measure-ment,as an anthropometric measurement is simplya construction of an observation or recording ofanattribute,whichcanbeaffectedbythemeasurers characteristics,methods u
12、sed in mea-surementandthemeasurementenvironment.Measurement consistency can be ascertained withrepeated measurements across time.With contin-ued repeated measurements,the concordant mea-surement results establish a measure of variabilitywithin measurement,measurer and methodologyapplied across time.
13、Systematic effects across thesevariables can be studied similarly.Some studies involving anthropometric mea-surements have investigated the topic of measure-ment consistency in relation to intrinsic qualities ofvariability within a given measurement.Gavan(1950)graded anthropometry dimensions in term
14、sof consistencies seen though expert anthropome-trists.Gavans work investigated the potentialsources of variance within a given measurement,and concluded that,consistency increased as:Thenumber of technicians decreased,the amount ofsubcutaneous tissue decreased,the experience ofthe technician increa
15、sed,and as the landmarkswere more clearly defined.(Gavan,1950)Otherworks looked into the consistencies within andbetweenmeasurers(GordonandBradtmiller,1992;Gordon et al.,1989;Ulijaszek and Mascie-Taylor,1994),and repeated measurement consis-tencies across groups of measurements(Relethford,1994).Thes
16、e have lead to increased awarenessaboutsourcesofmeasurementerrorinanthropometry,as well as acceptable levels ofreliability.In an attempt to improve measurement consis-tency within and across individuals who aremeasured,the postural conditions are usuallystandardized to a rigidly instructed,but propr
17、io-receptively imbalanced and erect posture.Studiesthat take anthropometric measurements in con-trolled postures or conditions remain benchmarksfor the fields of anthropometry and engineeringanthropometry(Dempster,1955;Dempster et al.,1959;Kroemer et al.,1997;Roebuck et al.,1975;Stoudt et al.,1965).
18、Other studies emphasized thestandard anthropometric position when studyingadults(Annis et al.,1991;Damon and Stout,1963;Molenbroek,1987),workers of all types(Intranontet al.,1988;Rempel and Serina,1995),andchildren(Donoso,1987;Sunnegardh et al.,1988).Measurement over clothing is generally notrecomme
19、nded due to a potential increase insystematic error(Roebuck et al.,1975),and mostlarge-scaleanthropometrystudieshavestrictguidelines regarding clothing(Gordon et al.,1989).Investigators who have evaluated the effectsof clothing on measurements found that measure-ments taken with clothing were system
20、aticallylarger than measurements made with no clothing(Paquette et al.,1999).Pett and Ogilive(1957)alsofound that clothing impactedcorrelations ofheight and weight.However,information aboutthe size and abilities of clothed people has usefulapplications.For example,studies have sought tomeasure perso
21、ns while in clothing to obtainmeasurements that were realistic in the setting towhichthemeasurementswouldbeapplied(AnthropologyResearchProject(Ed.),1978;Roebuck et al.,1975).Despite the multitudes of studies using andreporting anthropometric data,there are fewstudies that have reported concerns on c
22、apturinganthropometric data and the variable nature of itsresults.The 1988 US Army ANSUR study offeredguidelines for acceptable error estimates for aspecific set of measurement conditions that includenude individuals who were measured while in erectstandardized postures using conventional mea-sureme
23、nt approaches(Clauser et al.,1989).Addi-tional efforts are needed to assess the viability oferror checking mechanisms for different clothingconditions,postural considerations,and measure-ment approaches.Three-dimensional anthropometry has been inuse for well over a decade.Authors have describeddevic
24、es that have enabled this new measurementapproach to be used(Annis,1989;Brooke-Wavellet al.,1994;Coblentz et al.,1985;Hoekstra,1997).Methods may range from manual collection ofthree-dimensional locations of body landmarks viaelectromechanical probe or electromagnetic sen-sing systems to three-dimens
25、ional scanning ofentire body surfaces.To date,however,there hasARTICLE IN PRESSD.J.Feathers et al./International Journal of Industrial Ergonomics 33(2004)181190182been little published literature on the measurementconsistency of three-dimensional anthropometryapproaches.Nor has there been literature
26、 thatchronicles the differences and similarities betweentwo-and three-dimensional anthropometry mea-surement approaches and output data.Only a fewstudies have documented how three-dimensionaldata can be used in designing,although it appearsthat three-dimensional anthropometry has someimportant advan
27、tages over conventional measuresthat include,for example,data that can be morereadily used for three-dimensional human model-ing(Reed et al.,2000).The two studies reported here have investigatedsome potential sources of error in anthropometricmeasurement for conventional methods(i.e.,useof anthropom
28、eters and calipers)and an electro-mechanical method.Of particular interest weresources of error associated with clothing,and non-standardized seated postures.The research ques-tions included:1.How reliable are the three-dimensional methodsof data collection,as compared to the conven-tional approache
29、s that employ the use ofanthropometers and calipers?2.For the three-dimensional measurement meth-ods,will clothing lead to a systematic over-estimation of body dimension size or will it leadto greater variability in the measurements?3.What effects do non-standardized posturesassumed by users of whee
30、lchairs have on theconsistency of the anthropometric measure-ments obtained with conventional and electro-mechanical methods?2.Study 1:measurement consistency for acadaveric arm2.1.MethodsTwo participants with knowledge of anatomyand anthropometry laboratory experience per-formed repeated measures o
31、n a human cadavericarm(male,age 68).One of the participants was aphysical anthropologist with 8 years of formalcourseworkinmusculoskeletalanatomyandresearch experience in the measurement of osteo-logical features of the forearm using conventionalanthropometric measurement devices.The otherparticipan
32、t was an industrial engineer who hadcoursework in gross anatomy,and some experi-ence in the collection anthropometric data usingconventional anthropometric measurement toolssuch as anthropometers and calipers.Differencesbetween measurers in terms of reliability werehypothesized to decrease with the
33、repeated use ofthe three-dimensional approach since both parti-cipants had very little experience with the electro-mechanical device(three-dimensional approach).MeasurementsofBihumeralEpicondylarbreadth,Radiale-Stylionlength,andWristbreadth were made.These were chosen becausethey are typical measure
34、ments that are widelydistributed in terms of measurement consistency(Gavan,1950).Bihumeral breadth and Wristbreadth were thought to be obtained more easilyand more consistent than Radiale-Stylion length,since it requires locating body landmarks that wereless prominent.Three measurement approaches we
35、re used byeach of the measurers:Conventional,an electro-mechanicalapproachthatrequiredmanuallyidentifying the specific point of interest,and anelectromechanical approach that required scan-ning a body surface area.In the conventionalapproach,a sliding caliper was used to collect theBihumeral Epicond
36、ylar breadth and the Wristbreadth measurements,and an anthropometer tocollect the Radiale-Stylion length.The electrome-chanicaldeviceusedinthisstudywastheFaroArmt(Faro Technologies,Florida).Thedevice is an articulating arm with six degrees offreedom and a probe tip of 0.25in and a reportedprecision
37、of 0.3mm.With this device,the measureris required to manually manipulate the tip of theprobe to the desired body point or area,and thedevice records the three-dimensional coordinatesof the desired point(Fig.1).The first electro-mechanical approach required the measurer tomanipulate the probe tip to
38、the desired bodylocation,and the three-dimensional coordinates ofthe probe tip were recorded(point and clickmethod).In the second electromechanical condi-tion,the measurer was required to use the probe toARTICLE IN PRESSD.J.Feathers et al./International Journal of Industrial Ergonomics 33(2004)18119
39、0183manually scan a surface area that included thebody location of interest.For this condition anextreme point thought to represent the bodylocation(i.e.,most inferior,superior,distal orproximal point within the scan)was automaticallyidentifiedwiththeelectromechanicaldevice(scan method).The scan fun
40、ction was thoughtto automate the identification of an extreme pointof interest similar to the way a caliper is swept overa body region when measuring breadths.(Fig.2).Measurers were blinded to the outcomes of themeasurements to reduce learning effects associatedwith repeated measurements.The measure
41、mentmarks on the calipers and anthropometers werecovered throughout the measurement trials.Forthese measurements,the measurer adjusted thetool to the desired setting while remaining blindedto the value of the measurement and the experi-menter recorded the value and readjusted themeasurement device o
42、ff of the desired setting.Inboth electromechanical conditions,the measurerswere blinded to the three-dimensional coordinatesof the body points that were measured,as well asthe dimensions(e.g.,lengths and breadths)thatwere calculated from the three-dimensional co-ordinate data.The anthropometric meas
43、urements were calcu-lated as the point-to-point distances between thebody landmarks.Epicondylar breadth was definedas the distance between the Medial HumeralEpicondyle and the Lateral Humeral Epicondyle.Radiale-Stylion length was defined as the distancebetween the Radiale landmark of the Radius andt
44、he Radial-Styloid landmark of the Radius.Wristbreadth was defined as the distance between theRadial Styloid of the Radius and the Ulnar Styloidof the Ulna.Training was provided for all measurementapproaches,anatomical landmarks and anthropo-metric measurements.Trial measurements wereperformed to int
45、roduce the participants to themeasurement approaches,ensure the measurementtaskswerepracticed,andobtainabaselinemeasurementofeachmeasurersabilitywithrespect to the task.Repeated measurement sessionsARTICLE IN PRESSFig.1.The electromechanical arm used in this study had six degrees of freedom,a workin
46、g radius of 2m and accuracy of 0.3mm.Theprobe tip is shown here.Fig.2.An example of the area scanned with the probe and theextreme point that is digitized(Radial-Styloid)automatically.D.J.Feathers et al./International Journal of Industrial Ergonomics 33(2004)181190184were conducted to evaluate the e
47、ffects of measurer,measurementapproachandclothingonthereliability of the measurements.An overview session was first provided toorientate the participants to both conventionaland electromechanical approaches.The sessionopened with a brief overview of the study.Theparticipants were then asked to handl
48、e the instru-ments and take practice measurements of objectsin the laboratory for approximately 2h.This wasthe only time in which the participants saw themeasurement marks on the conventional tools.Fortheremainderofthestudy,instrumentmeasurement marks were covered to minimizeany learning effect that
49、 could occur within methodbecause of the repeated measurements.Two days after the orientation session,partici-pants returned to the laboratory and performed aninitial series of measurements for attaining baselineerror estimates of their respective performances.Two trials with repeated measures were
50、performed.First,a steel bar(FaroArm Certification bar,FaroTechnologies,Florida)with a rounded end wasmeasured to allow the participants to becomereacquainted with the measurement tools andmethods.In the second trial,Bihumeral Epicondy-lar breadth,Radiale-Stylion length,and Wristbreadth were made on