Home Email this page Print this page Bookmark this page Decrease font size Default font size Increase font size
Noise & Health  
 CURRENT ISSUE    PAST ISSUES    AHEAD OF PRINT    SEARCH   GET E-ALERTS    
 
 
Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
Email Alert *
Add to My List *
* Registration required (free)  
 


 
   Abstract
   Introduction and...
   Methods
   Results
   Discussion
   Recommendations
   Acknowledgments
   References
   Article Figures
 

 Article Access Statistics
    Viewed5236    
    Printed192    
    Emailed1    
    PDF Downloaded36    
    Comments [Add]    
    Cited by others 5    

Recommend this journal

 


 
  Table of Contents    
ARTICLE  
Year : 2011  |  Volume : 13  |  Issue : 51  |  Page : 99-112
The challenge of localizing vehicle backup alarms: Effects of passive and electronic hearing protectors, ambient noise level, and backup alarm spectral content

Auditory Systems Laboratory, Virginia Tech, Blacksburg, VA, USA

Click here for correspondence address and email
Date of Web Publication1-Mar-2011
 
  Abstract 

A human factors experiment employed a hemi-anechoic sound field in which listeners were required to localize a vehicular backup alarm warning signal (both a standard and a frequency-augmented alarm) in 360-degrees azimuth in pink noise of 60 dBA and 90 dBA. Measures of localization performance included: (1) percentage correct localization, (2) percentage of right--left localization errors, (3) percentage of front-rear localization errors, and (4) localization absolute deviation in degrees from the alarm's actual location. In summary, the data demonstrated that, with some exceptions, normal hearing listeners' ability to localize the backup alarm in 360-degrees azimuth did not improve when wearing augmented hearing protectors (including dichotic sound transmission earmuffs, flat attenuation earplugs, and level-dependent earplugs) as compared to when wearing conventional passive earmuffs or earplugs of the foam or flanged types. Exceptions were that in the 90 dBA pink noise, the flat attenuation earplug yielded significantly better accuracy than the polyurethane foam earplug and both the dichotic and the custom-made diotic electronic sound transmission earmuffs. However, the flat attenuation earplug showed no benefit over the standard pre-molded earplug, the arc earplug, and the passive earmuff. Confusions of front-rear alarm directions were most significant in the 90 dBA noise condition, wherein two types of triple-flanged earplugs exhibited significantly fewer front-rear confusions than either of the electronic muffs. On all measures, the diotic sound transmission earmuff resulted in the poorest localization of any of the protectors due to the fact that its single-microphone design did not enable interaural cues to be heard. Localization was consistently more degraded in the 90 dBA pink noise as compared with the relatively quiet condition of the 60 dBA pink noise. A frequency-augmented backup alarm, which incorporated 400 Hz and 4000 Hz components to exploit the benefits of interaural phase and intensity cues respectively, slightly but significantly improved localization compared with the standard, more narrow-bandwidth backup alarm, and these results have implications for the updating of backup alarm standards.

Keywords: Alarm spectrum, auditory warning signal localization, backup alarm, electronic earmuff, flat attenuation, hearing protector, musician′s earplug, reverse alarm, sound localization

How to cite this article:
Alali KA, Casali JG. The challenge of localizing vehicle backup alarms: Effects of passive and electronic hearing protectors, ambient noise level, and backup alarm spectral content. Noise Health 2011;13:99-112

How to cite this URL:
Alali KA, Casali JG. The challenge of localizing vehicle backup alarms: Effects of passive and electronic hearing protectors, ambient noise level, and backup alarm spectral content. Noise Health [serial online] 2011 [cited 2016 Dec 8];13:99-112. Available from: http://www.noiseandhealth.org/text.asp?2011/13/51/99/77202

  Introduction and Literature Background Top


One of the most common auditory warning signal on industrial and construction sites is the backup (i.e., reverse) alarm signal. The Occupational Safety and Health Administration (OSHA) regulations state that "No employer shall use any motor vehicle equipment having an obstructed view to the rear unless: (b)(4)(i) The vehicle has a reverse signal alarm audible above the surrounding noise level or: (b)(4)(ii) The vehicle is backed up only when an observer signals that it is safe to do so" (Part 1926.601[b][(4)]). [1] Although OSHA mandates installing audible backup alarms on construction motor vehicles, accidents that stem from vehicular backing maneuvers still occur frequently. Purswell and Purswell [2] investigated OSHA accident reports from approximately 1972 through 2001, and found that of the roughly 150 investigated backing accidents, approximately 43% occurred while the backup alarm was operable. This statistic raises a question about the effectiveness of backup alarms in occupational settings.

The purpose of industrial alarms, especially auditory ones, is to capture workers' attention about critical information. [3] To grab workers' attention, the auditory alarm must be recognizable (identifiable) as well as detectable. To initiate the reaction intended by the designers of the auditory signal, the signal should be informative, which connotes not only identification of the hazard but also location of it in many scenarios. The aforementioned OSHA regulation [1] emphasizes the signal detection aspect of the backup alarm signal design while ignoring the information-conveying aspect. As part of being informative, auditory warning signals need to be localizable in most applications, including for warning of an approaching vehicle. To be localizable, backup alarm signals should include in their frequency spectrum components that foster signal localization. In human hearings, frequencies below 1500 Hz and above 3000 Hz are important for localizing sounds in the horizontal plane outside the median plane. [4],[5] Also, to help localize horizontal sound sources in the median plane (i.e., front or back sound sources), the signal should have a broadband spectrum. Below 1500 Hz, the Interaural Time Difference (ITD), which is the difference in sound wave arrival time between the two ears, is used to localize sound sources in the horizontal plane outside the median plane. Above 3000 Hz, the Interaural Level Difference (ILD), which is the difference in sound wave level between the two ears, is used to localize sound sources in the horizontal plane outside the median plane. [4],[5] Most current commercial backup alarms emit near-pure tone signals, with a dominant frequency in the 1000-1400 Hz frequency range [6] and thus lack the frequencies that foster sound localization. In addition, the Society of Automotive Engineers (SAE) standard SAE J994b mandates a predominant frequency in the frequency range of 700-2800 Hz for backup alarms, which has limited coverage of localizable frequencies. [7] It is evident that commercially available backup alarms that consist of narrow-band spectra, which do not include the frequencies that enable interaural phase and intensity cuing, are not optimized; that is, they may not alert workers to the potential of hazardous situations resulting from vehicle reversing in part because their design, while being detectable, does not foster auditory localization. Furthermore, in the case of multiple vehicles in a construction environment, the rather obvious importance of knowing which of the vehicles is approaching the worker has been empirically verified in a field study conducted by Withington, in which 90% of the participants stated that it is important to know which of the surrounding vehicles is approaching them. [8] To accomplish this task, localization of the backup alarm is of obvious importance.

In the scientific literature, most of the investigations of backup alarm effectiveness have focused on the influence of hearing protection devices (HPDs), background noise levels, workers' hearing abilities, and the acoustic features of the backup alarm signal on its detection. [6],[9] A single prior study actually examined, on a subjective basis, the effect of the backup alarm's acoustic features on how workers perform in localizing the source of the signal. In this study, Withington [8] used self-report measures to assess workers' performance in localizing conventional backup alarms as well as broadband ones. The broadband alarm, consisting of a white noise-type signal, was reported by Withington to improve workers' localization performance over that reported for a conventional backup alarm. Although these results appear promising in terms of improving on-foot workers' safety, factors other than the acoustic features of backup alarms that may render workers' localization ability (e.g., background noise level and HPD-type) were not considered in the Withington study. Conventional passive HPDs are necessary as a countermeasure against noise-induced hearing loss in workers, but their use has been shown in a limited number of studies to be detrimental to sound localization in a horizontal plane. [10],[11],[12],[13],[14],[15],[16] As well as investigating the effect of the acoustic features of backup alarm signals on localizing them, it is also important to study the effect of different HPDs, including those that incorporate either passive or electronic augmentations (see Casali [17] for a review), to localize backup alarm signals. This should help industrial safety professionals select HPDs for their workers when localization is an important task and thus a comparison of a wide variety of passive and electronic HPDs was a major focus of the research discussed in this article. In addition, the research had the objective of determining the localization effects of broad-spectrum (i.e., pink) noise at the OSHA criterion level of 90 dBA and at a lower level of 60 dBA as well as the localization effects of a frequency-augmented backup alarm as compared with a standard backup alarm.


  Methods Top


Participants

A total of 12 participants, eight males and four females, all older than 18 years, participated as subjects in the localization experiment. All qualifying participants were tested with a Beltone Model 114 pure-tone audiometer and were found to have normal hearing (i.e., defined as ≤25 dBHL at 250, 500, 1000, 1500, 2000, 3000, 4000 and 6000 Hz in both ears) and a symmetry difference of ≤15 dBHL between the two ears. The symmetry requirement was implemented to avoid any possible hearing-related bias in localizing the study stimuli.

Experimental design

To assess localization performance with objective measurements, an 8 × 2 × 2 completely within-subject experimental design with three independent variables was applied [Figure 1]. The independent variables were the hearing protection condition (eight levels), background noise dB level (two levels), and type of backup alarm signal (two levels), detailed as follows:
Figure 1: Experimental design with assignment of participants to independent variables

Click here to view


HPD variable: The HPDs consisted of: (1) unoccluded (i.e., no HPD) condition, (2) Moldex Model-6604 foam earplug (SparkPlug TM earplug), (3) E-A-R Ultrafit TM pre-molded earplug, (4) E-A­-R HiFi TM earplug (a uniform or "flat" attenuation HPD), (5) E-A-R Arc TM earplug (a level-dependent HPD), (6) Bilsom Leightning TM L3HV passive earmuff, (7) Bilsom Impact TM dichotic sound transmission earmuff (having two unique microphones, each feeding the sound transmission circuit for one earcup to maintain true stereo hearing), [17] and (8) a custom-made diotic sound transmission earmuff (having one microphone that feeds the sound transmission circuit for both earcups, thereby negating any stereo cues). [17]

The rationale for selecting each HPD was as follows. The foam earplug was selected due to its popularity and prevalence in industry. In addition, the Moldex SparkPlug TM foam earplug has the highest noise reduction rating (NRR = 33) compared with the other HPDs under test. This was considered important because backup alarms are common in road construction environments where the exposure levels can be quite high, necessitating an HPD of high-NRR value. The E-A-R Ultrafit TM pre-molded earplug (NRR = 25) and the E-A-R HiFi TM earplug (NRR = 12) were selected based on several factors, including their ease of cleaning (i.e., hygienic benefit) and the fact that flanged polymer earplugs are popular in industry. Furthermore, unlike the attenuation profile of conventional passive earplugs that provide higher attenuation as frequency increases, the HiFi TM earplug (which has identical flange design to the Ultrafit TM ) has near-flat attenuation in the 100-8000 Hz frequency range. [17] This is believed to be beneficial to pitch perception and thus it was hypothesized that it potentially could provide an advantage in horizontal sound localization. The two-ended E-A-R Arc TM orifice-based, level-dependent earplug, an industrial version of the military Combat Arms TM earplug, was included because its level-dependent end provides minimal attenuation at frequencies below 1000 Hz under low-level noise conditions, and its attenuation automatically increases in impulsive noises above about 110-120 dB, such as those from a piledriver. [17] Thus, it was thought that the Arc TM earplug could have applications in certain industrial environments where intermittent noise having primarily impulsive components is prevalent (e.g., as in the construction industry). However, in this first localization investigation, impulsive noise was not investigated but, instead, the interest was in whether this earplug might show localization benefit during the "quiet" periods (i.e., 60 dBA noise level) that occur in intermittent noises due to its minimal attenuation at those levels. In addition, the lower attenuation at frequencies below 1000 Hz could provide listeners with the ITD sound localization cue, which may improve their localization performance. (For this experiment, only the level-dependent end of the Arc TM earplug was tested in view that the conventional end of the Arc TM is essentially the same as the Ultrafit TM earplug.)

The earmuffs, as a class of HPDs, were included for several reasons, including their prevalent use in certain construction environments as well as the fact that unlike earplugs, because they cover the pinnae, earmuffs remove any localization cueing that may be established by pinnae configuration. The inclusion of the Bilsom Leightning TM Hi-Visibility L3HV passive earmuff was based on its high attenuation (NRR = 30) relative to most other commercial earmuffs. Two active (electronic sound transmission) earmuffs, also known as sound restoration or sound pass-through devices, were also tested (see Casali [18] for a discussion of this class of electronic HPD). The custom-made diotic earmuff (NRR = 23 based on its passive mode of operation) consisted of a Bilsom Impact TM dichotic product, with the dichotic feature converted to a diotic design by the research team. This product was not a commercially available product at the time of this writing; however, other diotic designs have been available from various manufacturers. The Bilsom Impact TM dichotic earmuff (NRR = 23) was selected because this type of design is often considered by safety professionals for industrial environments that are characterized by intermittent noise. In relatively quiet environments, the Bilsom Impact TM dichotic muff amplifies the surrounding warning signals (and other sounds in its passband) after picking the signal through a microphone attached to the outside of each earcup. As noise increases above a certain level (i.e., 82 dBA as in the manufacturer's descriptions), sound transmission muffs of this type shut-off their pass-through circuits and the earmuff reverts to a passive mode to protect the listeners' hearing. Using a custom-made acoustical test fixture (per ISO/DIS 6290) with a 1-inch microphone connected to a Larson-Davis 3200D spectrum analyzer, an objective test was performed to determine the actual under-earcup output of two samples of the Bilsom Impact dichotic muff in noise levels widely spanning 82 dBA. In a noise field of surrounding, incident pink noise, presented at constant individual levels ranging from 75 dBA to 90 dBA, the Impact TM dichotic earmuff yielded under-earcup levels of approximately 90-91 dBA with the gain set to maximum (as it was set in the experiment described herein).

The difference between the diotic and the dichotic earmuffs in this study was limited to the design of how the microphones feed each earcup. In the diotic earmuffs, only one microphone attached to one of the earcups transmits the signal to both the earcups simultaneously. In the dichotic earmuffs, two microphones, one attached to each earcup, feeds each earcup separately. This difference in design was hypothesized to influence participants' localization performance as the diotic technology is expected to destroy the ITD and ILD while the dichotic one is expected to maintain them. Further discussion of these types of HPD technologies is provided by Casali. [17],[18]

Background noise level variable: The levels of background noise included in this study were 60 dBA and 90 dBA (i.e., "low" and "high," respectively) of pink noise (i.e., flat-by-octave within ±3 dB). The higher level, 90 dBA, was selected because it comprises the noise permissible exposure limit (PEL) (i.e., criterion level) of OSHA and because it is representative of many noise levels produced from various construction equipment as measured on construction sites. The 90 dBA level is 8 dB above the level at which the pass-through circuits of the electronic sound-transmission earmuffs were claimed by the manufacturer to shut-off (although the objective test described above yielded an under-earcup level of about 90-91 dBA), while 60 dBA is well below the shut-off level therefore invoking the amplification of the sound transmission devices. Also, it should be obvious that in construction sites, intermittent quiet periods do occur and workers are advised to keep their hearing protectors on, in view that unexpected noise can manifest at any time. Thus, this is the reason for including a 60 dBA level in the study, in that workers will still have to localize backup alarms during quiet periods, often with hearing protectors on.

Backup alarm signal variable: The third independent variable considered in this study was the type of the backup alarm signal. The participants' localization performance was assessed under both a standard backup alarm and a spectrally modified backup alarm. The "standard" backup alarm used includes dominant frequencies of 1000, 1250, and 3150 Hz while the spectrally modified backup alarm signal was augmented by the research team to include primary frequency components of 400 and 4000 Hz in addition to the same dominant frequencies of the standard alarm [Figure 2]. It was hypothesized that the modified backup alarm signal would provide more accurate sound localization judgments in the horizontal plane than would the standard backup alarm signal. This was because the additional frequencies (i.e., 400 and 4000 Hz) would enable use of the ILD and ITD in localizing the source of the alarm when it emanates longitudinally from outside the median plane. Additionally, the spectral cues in the modified signal (the broad bandwidth of the backup alarm signal) were expected to increase the performance of listeners with respect to the front-rear localization.
Figure 2: Standard and spectrally modified backup alarm spectra as used in the experiment

Click here to view


Experimental facility and apparatus

Background noise field: A hemi-anechoic room, with dimensions 19 ft long by 18 ft wide by 8.5 ft high in the Virginia Tech Auditory Systems Laboratory was used as the test environment. This established an approximate free-field over a reflecting plane to simulate an outdoor environment free of large obstructions and including a reflective ground surface, such as asphalt, concrete, or packed earth of a typical construction site. The 60 and 90 dBA background noise was generated by an  Atlas More Details Soundolier noise generator, Model GPN-1200A, routed through two Sony STR DE-135 amplifiers and two Phoenix Gold VSS2 speaker selectors, to four Infinity SM-155 loudspeakers [Figure 3]. To provide a flat-by-octave spectrum (i.e., pink noise) at the participant's head, two AudioControl C-131 one-third octave equalizers were used in the circuit. The spectrum of the background noise was verified before each experimental session, at the participant's head position, using a Larson-Davis 2800 Series Real-time Spectrum Analyzer and a Larson-Davis model 2559 ½-inch microphone, which was calibrated to 94 dBA at 1 kHz using a Quest QC-20 calibrator.
Figure 3: Experimental test environment with loudspeaker placement for pink noise and for backup alarms in 360-degrees azimuth. (An acoustically transparent curtain that was between the subject and the backup alarm loudspeakers is not shown)

Click here to view


Backup alarm sound field: The modified backup alarm signal was synthesized at a sampling rate of 192 KHz with 16-bit resolution using Adobe Audition TM software. The two additional frequencies, 400 and 4000 Hz, were added by using the software to add to the standard backup alarm signal. Using the same software, on each trial, the backup alarm signal under test was simulated to be approaching the participant's position by increasing its intensity over time. A DELL laptop computer was used to store the two backup alarm signals and input them to a Parasound P/LD-1100 Line Drive pre-amplifier and a Sony STR DE-135 amplifier. The backup alarm signal was routed through two Phoenix Gold VSS4 speaker selectors, enabling the alarm to be selective and on demand by the experimenter, output through one of eight Klipsch Comm Sat loudspeakers that were placed in a circle 6 ft away from the participant's head center position and at his/her ear height [Figure 3]. All loudspeakers were obscured from the participant's view by a black, acoustically transparent curtain.

Participant's localization response controls: To respond to each presented backup alarm signal, the participant used a stylus connected to a Fujitsu tablet PC to point, on a 360-degree digital compass with a 1-degree interval, to the location of the approaching backup alarm signal as he/she perceived it. A LabVIEW custom-made program installed on the tablet PC was used to present the 360-degree digital compass and an ENTER DIRECTION digital button. The participant was asked to press this button to store his/her response. To monitor the participant's use of the LabVIEW program, a 17" Dell monitor, connected to the video-out port of the tablet PC, was viewed by the experimenter. The participant did not know the number of backup alarm source loudspeakers and he/she was allowed to turn his/her head during each localization trial.

Procedure

Screening and familiarization session: During a screening and familiarization session, the participant read and signed an informed consent document and then underwent the audiometric test for normal hearing and bilateral symmetry. Also, during this session, the participant was instructed in the procedures of the experiment and practiced a familiarization task that required him/her to localize 16 approaching beeps (eight beeps presented from each Klipsch loudspeaker twice each) of the spectrally modified backup alarm signal in the lower background noise condition (60 dBA of pink noise) while being unoccluded. To meet the familiarization criterion, the participant must have been able to achieve at least 50% correct of his/her localization responses on this task.

Experimental sessions: There were four experimental sessions attended by each subject, and each session included factorial combinations of the independent variables [Figure 1] consisting of: two hearing protection conditions, both background noise levels, and both backup alarm conditions that were each presented twice. Presentation of both backup alarms was accomplished in a random fashion within a session, as were the background noise levels, to avoid order effects. Also, the order effect of assigning two hearing protection conditions to each of the four sessions for each participant was minimized via counterbalancing of order by a Latin Square. In each experimental session, participants performed the localization task in 64 experimental trials once per hearing protection condition. The 64 experimental trials consisted of presentation of two redundant trials for each of the two different backup alarm signals from each of the eight Klipsch speakers in the two background noise conditions. Before starting the experimental session, the participant was seated in a chair placed in the middle of the test room, refreshed on the experimental procedures, instructed to move his/her head if needed to localize the backup alarm signal, and fit properly with the assigned HPD by the experimenter. Next, the participant performed the 64 experimental trials of the localization task. In each experimental trial, the participant, wearing the assigned HPD (or open-ear condition) was presented with one of the background noise levels and one of the backup alarm types.

The backup alarm's sound level was increased during each trial to simulate a vehicle approaching from 240 ft away to 19 ft away from the listener. The rate of increase in intensity was calibrated to simulate an approach speed of 10 mph; this speed was selected because it represented one of the faster reversing speeds among construction vehicles. The change in sound level of the backup alarm at the participant's ear, corresponding to the simulated distance from the participant, is depicted in [Figure 4]. On each trial, once the backup alarm began to sound, the participant had 15 s to perceive the location of the backup alarm signal and 15 s to provide his/her estimate of the alarm's location using the tablet PCs' 360-degree compass. If the participant did not respond within 15 s, he/she was then instructed to make his/her immediate best guess for the perceived location of the backup alarm signal.
Figure 4: Backup alarm's simulated approach toward participant, with dBA increment by distance

Click here to view


Dependent measures of localization performance

Four objective, quantitative dependent measures were obtained from the data set to assess participants' localization performance: (1) percentage correct localization, i.e., any response within 22.5° to the right or left of the alarm's azimuthal location was considered correct; (2) percentage of right-left localization errors, i.e., an error occurred when an alarm that emanated from within an angle of ±45° from directly to the right of the subject was judged as coming from the left, and vice-versa; (3) percentage of front-rear localization errors, i.e., an error occurred when an alarm that emanated from within an angle of ±45° from directly in front of the subject was judged as coming from the rear, and vice-versa; and (4) localization absolute deviation in degrees from the alarm's azimuthal location, i.e., the error in judgment from the alarm's actual direction, which ranged from 0° to 180°.

Localization data reduction and statistical analysis

The raw data values that were collected via a computer using LabVIEW TM software from the 360-degree azimuth tablet PC compass pointer were reduced after exporting these data to a Microsoft Excel TM spreadsheet. In each session for each participant and for each noise level condition, the two redundant trials obtained for each of the two different backup alarm signals from each of the eight loudspeaker positions were averaged together. To investigate for statistically significant differences in the data set, a within-subject analysis of variance (ANOVA) was applied to each of the four dependent measures. An alfa level of 0.05 was used for all decisions regarding statistical significance; i.e., any differences reported herein were statistically significant at P ≤0.05. Main effects and interactions that were revealed by the ANOVAs as significant were further analyzed by Tukey's Honestly Significant Difference (HSD) test to determine the exact loci of significance, also at an alfa level of 0.05. The Statistical Analysis Software (SAS) program was used to compute the ANOVAs and the Statistical Minitab software was used to generate the mean and 95% confidence intervals' graphs depicted in the figures herein.


  Results Top


HPD main effects

For all four dependent measures, the ANOVAs revealed a significant main effect of HPD, as follows:

Percentage correct localization: The HPD effect on percentage of correct localization responses was significant, F (7,77) = 59.2, P <0.0001, with the means ranging from a low of 15.8% for the custom diotic earmuff to a high of 83.9% for the E-A-R HiFi TM flat attenuation earplug [Figure 5]. Post hoc analyses via Tukey's test indicated that the diotic earmuff (15.8% correct localization) significantly differed from all other listening conditions and that there were no differences between the unoccluded ear (82.2%), Moldex SparkPlug TM foam earplug (64.6%), E-A-R Ultrafit TM earplug (81.5%), E-A-R HiFi TM earplug (83.9%), E-A-R Arc TM earplug (79.9%), Bilsom Leightning TM earmuff (70.2%), and Bilsom Impact TM dichotic earmuff (66.3%).
Figure 5: Main effect of hearing protection devices on percentage correct localization; mean values shown on bars with 95% confidence intervals. Means with the same letter are not significantly different at P ≤0.05

Click here to view


Percentage of right-left localization errors: ANOVA also revealed a significant main effect of HPD on percentage of right-left localization errors, F (7,77) = 42.78, P <0.0001, with the means ranging from a low of 2.4% for the E-A-R HiFi TM earplug to a high of 36.5% for the custom diotic earmuff [Figure 6]. Post hoc analyses using Tukey's test indicated that there was a significant difference in the percentage of right-left localization errors only when comparing the custom diotic earmuff with all other HPD conditions; none of the other HPD conditions, including the open ear, differed significantly on right-left localization performance.
Figure 6: Main effect of hearing protection devices on percentage of right-left localization errors; mean values shown on bars with 95% confidence intervals. Means with the same letter are not significantly different at P ≤0.05

Click here to view


Percentage of front-rear localization errors: ANOVA showed a significant main effect of HPD, F (7,77) = 43.03, P <0.0001, with the means ranging from 4.0% for the E-A-R HiFi TM earplug to 43.8% for the diotic earmuff [Figure 7]. Post hoc analyses using Tukey's test indicated that there were no significant differences in front-rear localization when wearing any of the HPDs or when using the open ear, except for the diotic earmuff condition, which sharply degraded the front-rear localization.
Figure 7: Main effect of hearing protection devices on percentage of front-rear localization errors; mean values shown on bars with 95% confidence intervals. Means with the same letter are not significantly different at P ≤0.05

Click here to view


Localization absolute deviation in degrees: ANOVA also revealed a main effect of HPD on absolute deviation in degrees from the backup alarm's actual direction, at F (7,77) = 92.67, P <0.0001, with the means ranging from 11.8° for the E-A-R HiFi TM earplug to 83.2° for the diotic earmuff [Figure 8]. Post hoc analyses using Tukey's test again showed no differences between any of the HPDs or the open ear condition, except for the diotic earmuff, which caused significantly higher localization deviations than any other listening condition.
Figure 8: Main effect of hearing protection devices on absolute localization deviation in degrees; mean values shown on bars with 95% confidence intervals. Means with the same letter are not significantly different at P ≤0.05

Click here to view


Background noise level main effects

The background noise level was revealed via ANOVA to have a large magnitude, statistically significant main effect on all four dependent measures, and in all four cases, the post hoc tests confirmed this observation. For percentage correct responses, the means were 81.4% at 60 dBA and 54.7% at 90 dBA, with ANOVA statistics of F (1,11) = 103.7, P <0.0001. For percentage of right-left localization errors, the means were 5.4% at 60 dBA and 13.3% at 90 dBA, with ANOVA statistics of F (1,11) = 20.84, P = 0.0008. For percentage of front-rear localization errors, the means were 7.2% at 60 dBA and 17.3% at 90 dBA, with ANOVA statistics of F (1,11) = 28.31, P = 0.0002. For localization absolute deviation in degrees, the means were 17.5° at 60 dBA and 35.5° at 90 dBA, with ANOVA statistics of F (1,11) = 80.31, P <0.0001. From these results, it is evident that the 90 dBA-increased pink noise level, a level that is common to some industries and construction, results in much poorer localization than does a 60 dBA noise level. For three of the four dependent measures, localization performance was half as good (or less) in 90 dBA noise as compared with 60 dBA in this main effect of pink noise level.

HPD and background noise interaction effects

The ANOVA revealed a significant interaction between HPD and background noise level on percentage correct localization responses, F (7,77) = 9.51, P <0.0001, on percentage of front-rear errors, F (7,77) = 4.83, P = 0.0001, and on localization absolute deviation in degrees, F (7,77) = 7.27, P <0.0001. Specifically, in the 60 dBA background noise level, the means of percentage of correct localization responses ranged from a low of 14.8% for the diotic earmuff to a high of 98.9% for the unoccluded ear [Figure 9]. At 60 dBA, the Moldex foam earplug resulted in significantly poorer localization than any of the three E-A-R earplugs (Ultrafit TM , HiFi TM , and Arc TM ) and the open ear condition by approximately 20% or more. In the 90 dBA background noise level, the means for percentage correct localization ranged from a low of 16.7% for the diotic earmuff to a high of 71.1% for the E-A-R HiFi TM flat attenuation earplug, with the dichotic earmuff resulting in poorer localization than the E-A-R Ultrafit TM and HiFi TM earplugs and the Moldex foam earplug resulting in poorer localization than only the HiFi TM earplug [Figure 10].
Figure 9: Effect of hearing protection devices at 60 dBA background noise level (from interaction) on percentage correct localization; mean values shown on bars with 95% confidence intervals. Means with the same letter are not significantly different at P ≤0.05

Click here to view
Figure 10: Effect of hearing protection devices at 90 dBA background noise level (from interaction) on percentage correct localization; mean values shown on bars with 95% confidence intervals. Means with the same letter are not significantly different at P ≤0.05

Click here to view


In the 60 dBA pink noise, on the measure of percentage of front-rear localization errors, there were no differences among any of the HPD or open ear conditions, with means ranging from a low of 0% for the open ear condition to 9.7% for the Moldex foam earplug, with the exception of the diotic earmuff, which yielded 43.8% front-rear errors. On this same measure, in 90 dBA noise, all HPDs were associated with poorer localization than at 60 dBA, but the lack of significant differences among HPDs was similar to the 60 dBA results, except that the dichotic earmuff resulted in more front-rear errors than the E-A-R Ultrafit and HiFi TM earplugs [Figure 11]. Again, the diotic muff resulted in the worst front-rear localization performance at 90 dBA [Figure 11].
Figure 11: Effect of hearing protection devices at 90 dBA background noise level (from interaction) on percentage of front-rear localization errors; mean values shown on bars with 95% confidence intervals. Means with the same letter are not significantly different at P ≤0.05

Click here to view


On the measure of localization absolute deviation in degrees, in 60 dBA pink noise, the deviations ranged from a low of 3.4° in the open ear condition to a high of 83.8° in the diotic earmuff condition [Figure 12]. Again, on this measure, as occurred with the percent correct localization measure, at 60 dBA, the Moldex foam earplug was associated with higher localization deviations than either the open ear condition or any of the three E-A-R earplug conditions, as was the Bilsom passive earmuff, while the diotic earmuff had the highest localization deviations of all HPDs. When the noise level was raised to 90 dBA, the location deviations increased for all listening conditions, ranging from a low of 19.3° for the HiFi TM earplug to a high of 82.6° for the diotic earmuff [Figure 13]. At 90 dBA, only the Bilsom dichotic earmuff had significantly more localization deviations than the three E-A-R flanged earplugs, and all the other HPD conditions had similar localization deviations, except that again the diotic earmuff showed the worst localization performance.
Figure 12: Effect of hearing protection devices at 60 dBA background noise level (from interaction) on localization absolute deviation in degrees; mean values shown on bars with 95% confidence intervals. Means with the same letter are not significantly different at P ≤0.05

Click here to view
Figure 13: Effect of hearing protection devices at 90 dBA background noise level (from interaction) on localization absolute deviation in degrees; mean values shown on bars with 95% confidence intervals. Means with the same letter are not significantly different at P ≤0.05

Click here to view


Backup alarm signal main effects

The ANOVA tests revealed a significant main effect of backup alarm spectrum on three of four dependent measures; however, the magnitudes of the improvements shown by the spectrally modified alarm, although statistically significant, were numerically small. For percentage correct localization, the means were 66.1% for the standard alarm and 70.0% for the spectrally modified alarm, with ANOVA statistics of F (1,11) = 11.3, P = 0.0063, an improvement of 4% through high- and low-band frequency augmentation of the standard alarm. For percentage of right-left localization errors, the means were 11.3% for the standard alarm and 7.4% for the spectrally modified alarm, with ANOVA statistics of F (1,11) = 15.4, P = 0.0024, also an improvement of about 4%. For localization absolute deviation in degrees, the means were 28.9° for the standard alarm and 24.1° for the spectrally modified alarm, with ANOVA statistics of F (1,11) = 25.92, P = 0.0003, representing another improvement of almost 5%.

Noise level and backup alarm signal interaction effects

Using ANOVA, a significant interaction was found between background noise level and backup alarm signal type on percentage of right-left localization errors, F (1,11) = 18.02, P = 0.0014, on percentage of front-rear localization errors, F (1,11) = 5.79, P = 0.0349, and on localization absolute deviation in degrees, F (1,11) = 11.21, P = 0.0065. For all three measures, significant differences existed only between the means for the 60 dBA and the 90 dBA noise levels, as would be expected based on the main effect of the noise level discussed above. There were no differences between the standard and spectrally modified backup alarms' performance on any measure in the 60 dBA noise level. However, at 90 dBA, a trend of improvement due to the spectrally modified alarm was revealed, consisting of 6.8% fewer right-left errors, 3.5% fewer front-rear errors, and 8.3° lower localization deviation. More important, however, was the benefit of the spectrally modified alarm on directional errors as the noise level increased from 60 to 90 dBA. Front-rear localization errors for the standard alarm showed a significant increase of 12.0% as the noise level changed from 60 to 90 dBA but, with the spectrally modified alarm, the increase was non-significant at 8.1%. Right-left localization errors for the standard alarm showed a significant increase of 10.7% as the noise level changed from 60 to 90 dBA, but with the spectrally modified alarm, the increase was non-significant at 5.0%. Thus, there is a noise level-dependent advantage for the spectrally modified alarm.

HPD and backup alarm signal interaction effects

An ANOVA revealed a significant interaction between HPD and backup alarm type on percentage correct localization, F (7,77) = 2.2, P = 0.0432, on percentage of right-left localization errors, F (7,77) = 2.77, P = 0.0128, and on localization absolute deviation in degrees, F (7,77) = 3.31, P = 0.0039.

Percentage correct localization: A comparison of [Figure 14] (standard backup alarm) and [Figure 15] (spectrally modified backup alarm) shows that for all HPD conditions as well as the open-ear condition, the frequency-augmented backup alarm was associated with a trend of increased correct localization performance, although of small performance increments in some cases (i.e., HiFi TM and Arc TM earplugs, passive earmuff), the only exception being the diotic earmuff, which had very low performance in both alarm conditions. Improvements in localization performance were as high as 10.2%, as occurred with the electronic dichotic earmuff.
Figure 14: Effect of hearing protection devices with the standard backup alarm (from interaction) on percentage correct localization; mean values shown on bars with 95% confidence intervals. Means with the same letter are not significantly different at P ≤0.05

Click here to view
Figure 15: Effect of hearing protection devices with the spectrally modified backup alarm (from interaction) on percentage correct localization; mean values shown on bars with 95% confidence intervals. Means with the same letter are not significantly different at P ≤0.05

Click here to view


Percentage of right-left localization errors: Corroborating the results of the percent correct localization-dependent measure, the percentage of right-left localization errors also demonstrates consistent trends in improvement for most HPD conditions when participants localized the spectrally modified backup alarm as compared with the standard alarm. It is evident in a comparison of [Figure 16] (standard backup alarm) and [Figure 17] (spectrally modified backup alarm) that for all HPD conditions, with the exception of the HiFi TM earplug (which showed a slight degradation), the use of the spectrally modified alarm was associated with fewer right-left localization errors. The largest improvement resulted for the Moldex foam earplug, at 9.7% fewer errors. Not counting the diotic earmuff, other improvements varied from 0.3% for the Arc TM earplug to 6.6% for the dichotic earmuff.
Figure 16: Effect of hearing protection devices with the standard backup alarm (from interaction) on percentage of right-left localization errors; mean values shown on bars with 95% confidence intervals. Means with the same letter are not significantly different at P ≤0.05

Click here to view
Figure 17: Effect of hearing protection devices with the spectrally modified backup alarm (from interaction) on percentage of right-left localization errors; mean values shown on bars with 95% confidence intervals. Means with the same letter are not significantly different at P ≤0.05

Click here to view


Localization absolute deviation in degrees: Again, in agreement with the two prior dependent measures, the absolute deviation measure reveals the consistent trends in localization improvements with most HPDs as well as in the open-ear condition when the spectrally modified backup alarm was used. Comparing [Figure 18] (standard backup alarm) and [Figure 19] (spectrally modified backup alarm), all HPD conditions, as well as the open ear, exhibit a trend toward smaller localization deviations under the use of the spectrally modified alarm. While improvements were quite small for the HiFi TM earplug at 0.7°, Arc TM earplug at 1.3°, and passive earmuff at 2.9°, other HPDs showed substantial improvements, including the Moldex foam earplug at 10.4° and the dichotic earmuff at 10.7°.
Figure 18: Effect of hearing protection devices with the standard backup alarm (from interaction) on localization absolute deviation in degrees; mean values shown on bars with 95% confidence intervals. Means with the same letter are not significantly different at P ≤0.05

Click here to view
Figure 19: Effect of hearing protection devices with the spectrally modified backup alarm (from interaction) on localization absolute deviation in degrees; mean values shown on bars with 95% confidence intervals. Means with the same letter are not significantly different at P ≤0.05

Click here to view



  Discussion Top


Conclusions regarding HPDs

HPDs and localization accuracy: This experiment has demonstrated that if on-ground workers need to rely on backup alarms to localize approaching vehicles (and such alarms certainly are important as a safety countermeasure on many industrial and construction sites), the choice of a hearing protector can influence that localization. While it is the case that the graphs of main effects of HPDs [Figure 5], [Figure 6], [Figure 7] and [Figure 8] on each dependent measure of localization did not exhibit particular advantages for the open-ear condition, any of the four passive earplugs, or either the passive or the dichotic earmuff, there were HPD-specific differences within the various interaction effects. In the lower noise level (60 dBA), the high-attenuation Moldex SparkPlug TM foam earplug resulted in a lower percentage correct localization than any of the three lower attenuation E-A-R-flanged earplugs and also lower correct localization than the open-ear condition [Figure 9]. This same trend for poorer performance with the Moldex foam earplug, as well as the Bilsom passive earmuff, also occurred for the localization absolute deviation measure in 60 dBA noise (but not in 90 dBA noise) [Figure 12] and [Figure 13]. When the noise level was at 90 dBA [Figure 10], only the E-A-R HiFi TM flat attenuation earplug showed an advantage in percentage correct localization, and this advantage was only over the Moldex foam earplug and the Bilsom dichotic earmuff. Furthermore, there was no significant improvement of the HiFi TM earplug over the standard pre-molded earplug, the E-A-R UltraFit TM , which incorporates the exact same flanged seal design as the uniform attenuation earplug.

Also notable was that in the high-noise level condition (90 dBA), none of the four passive earplugs showed any disadvantage in localization accuracy when compared with the open ear. Interestingly, while the dichotic earmuff provided 31% better percent correct localization than the diotic muff in 90 dBA noise, it did not have any significant advantage over the localization performance achieved when the participant wore the conventional Bilsom passive earmuff and, furthermore, the dichotic earmuff demonstrated 18.2% and 23.4% poorer percent correct localization performance compared with the Ultrafit TM and HiFi TM earplugs, respectively [Figure 10]. This is possibly due to the fact that the pinnae cues are largely lost due to the earmuff's non-directional cups, and this does not occur with the earplugs that do not obscure the pinnae. Finally, it is important to note that the custom diotic earmuff consistently ranked as the worst HPD for localization on all four dependent measures by a large margin.

HPDs and directionality errors: A discussion of the effects on directional backup alarm judgment confusions (i.e., front-rear and right-left) is also needed. For the main effects of HPD on these two dependent measures, there were no differences among any of the HPD conditions or the open-ear condition, with the exception of the diotic earmuff exhibiting the worst performance on both directional error measures [Figure 6] and [Figure 7]. However, the interaction effects shed more light on directional confusions and how they are influenced by specific HPDs. For instance, confusions of front-rear alarm directions were most significant in the 90 dBA noise condition, where the UltraFit TM and HiFi TM earplugs exhibited significantly fewer front-rear confusions than either of the electronic muffs [Figure 11]. It is important to note that the dichotic earmuff, for which the sound transmission circuit would have shut-off in the 90 dBA noise level , showed a disadvantage in front-rear confusions (as well as in localization absolute deviation) when compared with the UltraFit TM earplug (by 12.2%) and the HiFi TM earplug (by 14.3%) at this noise level, but that these differences did not manifest at the 60 dBA level, at which the sound transmission circuit of the dichotic earmuff was operating and amplifying. This finding attests to the fact that the human pinnae configuration, which affords front-rear acoustical cue discrimination, is at least partially preserved with some earplugs, but is obscured by earmuffs. In regard to front-rear errors, as limited to among the four earplugs, there were no differences at 60 dBA, and although clearly not statistically significant at 90 dBA [Figure 11], the flat attenuation HiFi TM earplug did have the numerically lowest confusions at 7.6% compared with a high of 15.6% for the conventional SparkPlug TM foam earplug (and to 21.9% for the dichotic earmuff). In regard to right-left errors (averaged across both backup alarms and both noise levels), there were no significant differences between any of the HPDs, except for the diotic earmuff, which was associated with 37% right-left confusions compared with a low of 2% for the HiFi TM earplug to a high of 9% for the Moldex foam earplug [Figure 6].

Diotic vs. dichotic earmuffs and localization: A prominent result, consistent through all four dependent measures, was that the custom diotic earmuff significantly degraded participants' localization performance. The difference in localization performance between the diotic earmuff and the next poorest-performing HPD (which in most cases was the Moldex foam earplug) was 49% in percentage correct localization, 28% in right-left localization errors, 31% in front-rear localization errors, and 55° in localization absolute deviation from the actual direction of the backup alarm signal. It is evident from these results that the single microphone design of the diotic earmuff (which was otherwise identical to the dichotic earmuff) renders binaural auditory localization cues (i.e., both those from low-frequency ITD and high-frequency ILD) unusable to the ears, since the single microphone feeds the same acoustic signal to both ears. The results, which were consistent across all localization measures, offer clear evidence that diotic sound transmission design in earmuffs negates any advantages of binaural hearing in localization tasks. Although less expensive than dichotic designs due to their single microphone/amplifier system, diotic designs should be avoided based on these results, which are in line with the findings of previous studies. [10]

On all four measures in the main effects and interactions, the electronic earmuff of dichotic design offered no localization benefit over passive earmuffs of the same structural design (without electronics); thus, their added cost and maintenance requirements cannot be justified, at least for workers with normal hearing, as in this experiment. Of course, it must be recognized that this study only investigated one model of sound transmission earmuff and thus the results are not generalizable to other dichotic devices on the market, which vary in design features (see Casali, in press, for a review). [18]

Based on the composite findings of this localization study, it does appear that for at least the 90 dBA condition of this study, the uniform attenuation qualities of a moderate-attenuation HiFi TM earplug offer some advantage over certain other earplugs , as well as over electronic sound-transmission earmuffs, in maintaining localization to a backup alarm; however, it must be recognized that this is based on an experiment with specific conditions and thus the results may not be generalizable to other signals, noises, or HPDs.

It is important to recognize that certain passive HPDs, and most typically the passive earplugs with the exception of the high-attenuation Moldex foam earplug, showed no disadvantages in localization performance in most noise and backup alarm scenarios when compared with the unoccluded ear condition. While these results are clearly situation- and HPD-specific, the lack of a passive HPD-induced disadvantage to localization refutes the oft-given excuse by workers to remove their hearing protectors when in the vicinity of backup alarm signals, at least for normal hearing individuals, as were the participants in this study.

Conclusions regarding background noise level

Pink noise, presented at 60 dBA and 90 dBA, was selected for its qualities as a broad-spectrum, flat-by­-octaves noise that can, with some limitations, be generalized as a reasonable in-lab simulation of broadband noises found in some industries and construction. This type of broadband noise is a strong masker of narrowband backup alarm frequencies due both to direct frequency overlap masking as well as the upward spread of masking, which occurs at high sound levels (i.e., at the 90 dBA level in this study). The pink noise effect on localization was indeed evident in the results of this study, with strong main effects of background noise level, wherein all four dependent measures of localization were degraded when the background noise level was increased from 60 to 90 dBA. There were approximately 27% fewer correct localization responses in 90 dBA noise than in 60 dBA, and only about half of the localizations were correctly identified in 90 dBA. Also, right-left and front-rear errors, in addition to the localization absolute deviation from the backup alarm direction, more than doubled in percentage when background noise level was increased from 60 dBA to 90 dBA. From these data, it is evident that in high-noise levels, more attention needs to be devoted by safety professionals to ensure that workers are attentive to, and concentrate on, detecting and orienting to backup alarms. Not only are alarms more likely to be missed in high-noise levels, they are also much less likely to be localized correctly as compared with moderate noise conditions.

Conclusions regarding backup alarm signal type

The main effects of the type of backup alarm signal were consistent, and demonstrated a statistically significant advantage of the spectrally modified alarm over the standard alarm on three of four measures: percentage correct localization, percentage of right-left localization errors, and absolute deviation from the actual direction of the backup alarm signal. While the magnitudes of the localization improvements produced by the spectrally modified alarm were small (i.e., about 4-5%), their consistency and statistical significance warrants additional attention to the apparent benefits. That is, the benefit of adding 400 and 4000 Hz to the standard backup alarm signal (which peaks at about 1000, 1250 and 3150 Hz) is not only evidenced in the main effects, it furthermore is apparent in the interaction between the backup alarm signal and the background noise level. When the background noise level was increased from 60 to 90 dBA and the spectrally modified alarm was used, the participants were more resistant to right-left and front-rear localization errors than when the standard alarm was used. These metrics showed statistically significant degradations with the standard alarm as the noise level increased, but that was not the case with the modified alarm. While the percentage reduction of directional errors with the spectrally modified alarm was numerically small, it was indeed statistically significant, and in the case of right-left confusions, the change in errors between 60 and 90 dBA conditions was about half that, in percentage, of the standard alarm. This is important because many workplaces have noise characteristic of the 90 dBA broadband noise used in this study.

It is also important to note that the benefits of the spectrally modified alarm were HPD-specific; i.e., localization performance under certain (but not all) HPDs significantly improved when the modified alarm was used as compared with the standard alarm [Figure 14], [Figure 15], [Figure 16], [Figure 17], [Figure 18], [Figure 19]. This effect was generally consistent, and the HPDs that tended to realize the largest benefits of the spectrally modified alarm were the high-attenuation Moldex foam earplug and the electronic Bilsom dichotic earmuff. Improvements in percentage correct localization and reductions in right-left errors were as high as about 10%, and reductions of localization absolute deviation were as high as about 11°. These small but significant improvements in localization performance on an HPD-specific basis give rise to the need for more research on further improvements, both spectral and otherwise, to backup alarms as well as further research on the actual design characteristics of HPDs, which may influence alarm detection and localization.


  Recommendations Top


The research was comprised of a complex factorial experiment with four dependent measures of localization performance. Based on the statistical analyses of the resultant data, it is clear that there are differences in normal-hearing humans' localization performance that depends on HPD type and design, background noise level, and backup alarm spectral content.

More research on hearing protection design is needed; e.g. such that dichotic sound transmission (i.e., signal pass-through) HPD designs result in better localization (as well as detection), prior research in hearing science indicates that binaural hearing is advantageous in localizing signals and, based on the results of this study, it is clear that the diotic HPD design, which eliminates important interaural cues, should not be used in hearing protectors of the sound-transmission variety. Furthermore, certain passive HPD designs, such as level-dependent devices and flat attenuation earplugs, which show promise for improved hearing of alarms, may be subject to further optimization such that alarm localization is fostered. Such improvements could improve the situational awareness that construction and industrial workers need in order to evade vehicles in their work environments. The hearing sense is so important in this regard, in that, unlike vision, it is omnidirectional and does not have to be focused in the direction of a warning signal. However, care must be taken to provide appropriate hearing protection and alarms, which take advantage of binaural hearing cues in both phase (ITD) and intensity (ILD) differences that occur between the ears. Based on the composite findings of this localization study, it does appear that the attenuation qualities of a moderate- and flat-attenuation earplug has certain advantages in very specific instances over some, but not all other earplugs, as well as over electronic sound-transmission earmuffs, in providing auditory localization to a backup alarm; however, it must be recognized that this conclusion is based on an experiment with specific conditions and thus the results may not be generalizable to other signals, noises, or HPDs.

It is important to recognize that certain passive HPDs, and most typically the passive earplugs with the exception of the high-attenuation Moldex foam earplug, showed no disadvantages in localization performance in most noise and backup alarm scenarios when compared with the unoccluded ear condition. While again these results are clearly situation- and HPD-specific, the lack of a passive HPD-induced disadvantage to localization does not give credence to the oft-given excuse by workers to remove their hearing protectors when in the vicinity of backup alarm signals, at least for normal hearing individuals, as were the participants in this study.

Finally, this research provided important evidence that the backup alarm signal can be simply and easily improved by spectral modification, yielding small but significant improvements in localization performance. This research intentionally modified only one parameter of the backup alarm, adding 400 and 4000 Hz to the standard backup alarm spectrum. Given the improvements resulting from this spectral change, further research is now needed on the spectral details of backup alarms as well as possible changes in the temporal domain (e.g., the "beep rate"), or even other acoustic parameters.


  Acknowledgments Top


This research was supported by a construction safety grant from the National Institute for Occupational Safety and Health (NIOSH). The conclusions herein are those of the authors, and are not intended to represent the position of NIOSH.

 
  References Top

1.Occupational Safety and Health Administration (OSHA). Occupational safety and health standards: Motor vehicles (29 CFR, Part 1926.601). Washington, DC: Office of the Federal Register; 2000.   Back to cited text no. 1
    
2.Purswell JP, Purswell JL. The effectiveness of audible backup alarms as indicated by OSHA accident investigation records. In: Bittner AC, Champney PC, Morrissey SJ, editors. Advances in Occupational Ergonomics and Safety. Amsterdam, Netherlands. : IOS Press; 2001. p. 444-50.   Back to cited text no. 2
    
3.Endsley MR, Bolte B, Jones DG. Designing for situation awareness: An approach to user-centered design. Boca Raton (FL): CRC press; 2003.   Back to cited text no. 3
    
4.Hartmann WM. How we localize sound. Phys Today 1999;52:24-9.   Back to cited text no. 4
    
5.Middlebrooks JC, Green DM. Sound localization by human listeners. Annu Rev Psychol 1991;42:135-59.   Back to cited text no. 5
    
6.Laroche C, Lefebvre L. Determination of optimal acoustic features for reverse alarms: Field measurements and the design of a sound propagation model. Ergonomics 1998;41:1203-21.   Back to cited text no. 6
    
7.Society of Automotive Engineers (SAE). Performance, Test, and Application Criteria for Electronically Operated Backup Alarm Devices (ANSI/SAE J994b-1978). Warrendale, PA; 1978.   Back to cited text no. 7
    
8.Withington DJ. Reversing goes broadband. Quarry Manage 2004;31:27-34.   Back to cited text no. 8
    
9.Casali JG, Robinson GS, Dabney EC, Gauger D. Effect of electronic ANR and conventional hearing protectors on vehicle backup alarm detection in noise. Hum Factors 2004;46:1-10.   Back to cited text no. 9
    
10.Noble W, Murray N, Waugh R. The effect of various hearing protectors on sound localization in the horizontal and vertical planes. Am Ind Hyg Assoc J 1990;51:370-7.   Back to cited text no. 10
    
11.Atherley GR, Noble WG. Effect of ear-defenders (ear-muffs) on the localization of sound. Br J Ind Med 1970;27:260-5.   Back to cited text no. 11
    
12.Atherley GR, Else D. Effect of ear-muffs on the localization of sound under reverberant conditions. Proc R Soc Med 1971;64:203-5.   Back to cited text no. 12
    
13.Abel SM, Hay VH. Sound localization. The interaction of aging, hearing loss and hearing protection. Scand Audiol 1996;25:3-12.   Back to cited text no. 13
    
14.Noble WG, Russell G. Theoretical and practical implications of the effects of hearing protection devices on localization ability. Acta Otolaryngol 1972;74:29-36.   Back to cited text no. 14
    
15.Simpson BD, Bolia RS, McKinley RL, Brungart DS. The impact of hearing protection on sound localization and orienting behavior. Hum Factors 2005;47:188-98.   Back to cited text no. 15
    
16.Bolia RS, D'Angelo WR, Mishler PJ, Morris LJ. Effects of hearing protectors on auditory localization in azimuth and elevation. Hum Factors 2001;43:122-8.   Back to cited text no. 16
    
17.Casali JG. Advancements in hearing protection: Technology, applications, and challenges for performance testing and product labeling. Proceedings of the International Congress and Exposition on Noise Control Engineering; Rio de Janeiro, Brazil: 2005. p. 2097-118.   Back to cited text no. 17
    
18.Casali JG. Electronic augmentations in hearing protection technology circa 2009 including active noise reduction, electronically-modulated sound transmission, and tactical communications devices: Review of design, testing, and research. Int J acoustics & Vibration 15:(4):168-86.  Back to cited text no. 18
    

Top
Correspondence Address:
John G Casali
Grado Department of Industrial and Systems Engineering, 519G, Whittemore Hall, ISE Department, Virginia Tech University, Blacksburg, VA 24061
USA
Login to access the Email id

Source of Support: Construction safety grant from the National Institute for Occupational Safety and Health (NIOSH), Conflict of Interest: None


DOI: 10.4103/1463-1741.77202

Rights and Permissions


    Figures

  [Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7], [Figure 8], [Figure 9], [Figure 10], [Figure 11], [Figure 12], [Figure 13], [Figure 14], [Figure 15], [Figure 16], [Figure 17], [Figure 18], [Figure 19]

This article has been cited by
1 Augmented warning sound detection for hearing protectors
Eric R. Bernstein,Anthony J. Brammer,Gongqiang Yu
The Journal of the Acoustical Society of America. 2014; 135(1): EL29
[Pubmed] | [DOI]
2 Azimuthal auditory localization of gunshots in a realistic field environment: Effects of open-ear versus hearing protection-enhancement devices (HPEDs), military vehicle noise, and hearing impairment
Talcott, K.A., Casali, J.G., Keady, J.P., Killion, M.C.
International Journal of Audiology. 2012; 51(1): 20-30
[Pubmed]
3 Azimuthal auditory localization of gunshots in a realistic field environment: Effects of open-ear versus hearing protection-enhancement devices (HPEDs), military vehicle noise, and hearing impairment
Kristen A. Talcott,John G. Casali,John P. Keady,Mead C. Killion
International Journal of Audiology. 2012; 51(S1): S20
[Pubmed] | [DOI]
4 Detection of alarms and warning signals on an digital in-ear device
Marc-André Carbonneau,Narimene Lezzoum,Jérémie Voix,Ghyslain Gagnon
International Journal of Industrial Ergonomics. 2012;
[Pubmed] | [DOI]
5 Advanced hearing protection and communication: Progress and challenges
Giguére, C., Laroche, C., Brammer, A.J., Vaillancourt, V., Yu, G.
Proceedings of the Institute of Acoustics. 2011; 33 1((PART 3)): 225-233
[Pubmed]



 

Top