Side effects of level dependent hearing protectors
Konrad Rundqvist
Industrial Design Engineering, master's level 2020
Luleå University of Technology
Department of Business Administration, Technology and Social Sciences
Side effects of level dependent hearing protectors
AUTHOR: Konrad Rundquist 2020
SUPERVISOR: Örjan Johansson REVIEWER: First name Surname
EXAMINER: Lena Abrahamsson
MSc in INDUSTRIAL DESIGN ENGINEERING
Department of Business Administration, Technology and Social Sciences CIVILINGENJÖR I TEKNISK DESIGN
Master of Science Thesis in Industrial Design Engineering Side Effects of Safety equipment
Hearing protection devices effects on sound identification
© Konrad Rundquist
Published and distributed by Luleå University of Technology SE-971 87 Luleå, Sweden Telephone: + 46 (0) 920 49 00 00
Printed in Luleå Sweden by
Luleå University of Technology Reproservice Luleå, 2020
Acknowledgement
I would like to thank my supervisor, Örjan Johansson, for providing me with guidance and excitement during this project. I am also very grateful for having the chance to work, not only with 3M Peltor, but with Oscar Kårekull and Magnus Johansson who have taught me a lot about hearing protectors, the human ear, acoustics in general, and most importantly, what it is like to work as an engineer.
I don’t think this thesis would exist if it weren’t for friends and family. My mother and father have always shown interest in what I do and even provided with some unrecognized input to this project. The male choir Snapsakademien have for six years provided me with tremendous amounts of laughter every Wednesday at rehearsals, making the time in Lulea fly by.
This thesis started with me moving to Linköping, so I would like to take the chance to thank my wonderful girlfriend for embracing me into her heart and home.
Abstract
Hearing protectors are used in noisy environments to attenuate damaging sound levels. Problems are reported to arise from using hearing protectors, since useful sounds also are attenuated. Not only are sound levels attenuated, but other human abilities are affected by hearing protectors, such as the ability to identify and localize sounds. Manufacturers of hearing protectors also develop level-dependent hearing protectors that are supposed to create a better listening experience for the user than does passive hearing protectors. This thesis investigates how the ability to identify sound is affected by hearing protectors in the Swedish processing industry and to what extent level-dependent hearing protectors improve the user’s ability to identify sound. Semi-structured interviews were conducted by asking questions based on the projects research questions and theories found from a literary study. A questionnaire was sent out with similar questions for verification purposes. The current literature shows that the human ability to identify sound deteriorates by the occlusion from hearing protectors. Indications are such that hearing protectors with a level- dependent function deteriorate the ability less than does passive protectors. The relevance of this in the processing industry does not seem to be very palpable. Users indicate that it would be dangerous for this ability to be deteriorated in the workplace and that level-dependent hearing protectors does a better job a preserving sound identification than passive protectors, but their sound identification ability is not deteriorated enough by passive hearing protectors for problems to arise.
KEYWORDS: Level-dependent hearing protectors, method study, unconscious behavior, sound identification
Sammanfattning
Hörselskydd används in bullriga miljöer för att dämpa skadliga ljudnivåer.
Hypotesen är att användandet av hörselskydd kan ge ökade risker och påverka beteendet negativt hos användaren. Inte bara försvagas ljudnivåer, men andra mänskliga förmågor påverkas av hörselskydd, såsom förmågan att identifiera och lokalisera ljud. 3M Peltor utvecklar nivåberoende hörselskydd som ska skapa en bättre lyssningsupplevelse för användaren än passiva hörselskydd. I detta examensarbete undersöks hur förmågan att identifiera ljud påverkas av hörselskydd inom den svenska processindustrin och i vilken omfattning nivåberoende hörselskydd anses bättre för användaren. Semi-strukturerade intervjuer genomfördes genom att ställa frågor utifrån projektets frågeställningar och teorier från en litteraturstudie. En enkät sändes ut med liknande frågor i syfte att verifiera resultaten från intervjuerna. Den aktuella litteraturen visar att den mänskliga förmågan att identifiera ljud försämras av ocklusion från hörselskydd. Tidigare genomförda undersökningar indikerar att nivåberoende hörselskydd ger något bättre riktiningshörande än passiva skydd. Relevansen av detta i processindustrin verkar däremot inte vara särskilt påtaglig. Från de fåtal kvalitativa intervjuer som genomfördes under studien, uppger användarna att det skulle vara farligt att försämra denna förmåga på arbetsplatsen och att nivåberoende hörselskydd gör ett bättre jobb med att bevara ljudidentifieringen än passiva skydd. Deras förmåga till ljudidentifiering försämras dock inte tillräckligt av passiva hörselskydd för att problem ska uppstå.
NYCKELORD: Nivåberoende hörselskydd, metodstudie, omedvetet beteende, ljudidentifiering
Content
INTRODUCTION 3
1.1 Background 3
1.2 Stakeholders 3
1.3 Objective and aims 4
1.4 Project scope 4
1.5 Thesis outline 4
2 CONTEXT 5
2.1 Current state 5
2.1.1 3M Peltor 5
2.1.2 The Swedish Work Environment Authority 6 2.2 Level-dependent hearing protectors 7
3 THEORETICAL FRAMEWORK 9
3.1 Industrial Design Engineering 9
3.2 Accidents and inconvenience in the workplace 14
3.3 Work Environment 16
3.4 Data collection 16
3.4.1 Interviews 16
3.4.2 Polls 17
3.5 The AUDITORY system 9
3.6 Occlusion of auditory system 11
3.6.1 Sound localization 13
3.7 Unconscious effects from sound environments 17
4 METHOD AND IMPLEMENTATION19
4.1 Process 19
4.2 Project planning 19
4.3 Methods 20
4.3.1 Literature Review 20
4.3.2 Interviews 20
4.3.3 Constructing the interview 20 4.3.4 Constructing the questionnaire 23
4.4 Implementation 23
4.4.1 Data analysis 23
4.4.2 Collecting quantitative data 25
4.5 Method discussion 26
4.5.1 Interviews 26
4.5.2 Polls 26
5 RESULTS 27
5.1 Results of literature study 27 5.2 results of Context Immersion 27
5.2.1 Data analysis 28
5.2.2 Interviews 28
5.2.3 Polls 33
6 DISCUSSION 36
6.1 Sound identification 36
6.2 The method study 38
6.3 Relevance 39
6.4 Conclusions 39
6.4.1 Project objective and aims 39
6.4.2 Research question 1 40
6.4.3 Research question 2 40
6.4.4 Research question 3 41
6.4.5 Research question 4 41
6.4.6 Research question 5 42
6.5 RECOMMENDATIONS 42
7 REFERENCES 44
List of appendixes
Appendix 1. Interview questions 1 page
Appendix 2. Swedish translation of sound descriptions 1 page Appendix 3. Questionnaire (Swedish) 3 pages
List of figures
Figure 1. Work-related issues, 2018 ... 6 Figure 2. Linear attenuation effect from passive HPs... 7 Figure 3. Illustration of how the amplification is lower for higher SPLs in LDHPs. . 8 Figure 4. Accidents in the workplace, 2018. ... 14 Figure 5. REAT ON/OFF for 3M™ 3M PELTOR™ WS Alert XPI ... 12 Figure 6. The Speech spectrums main area.
A = Scope of fundamental tones. B =Vowels. C = Voicing consonants
(e.g. m, n and v). D = Voiceless consonants (e.g. s, t and k) ... 13 Figure 7 The Gantt scheme used in the project ... 19 Figure 8. The process of coding the transcript ... 24
List of words
Word Abbreviation
Level dependent function LDF Level dependent hearing
protectors LDHP
A-weight decibel accounting for the loudness perceived by humans
dB(A)
Hearing protection
device HP
Sound pressure level SPL
Introduction
The Swedish industry has a strong tradition of minimizing risk in the workplace.
Efforts are continually made in order to create a safer work environment. One of these efforts is the usage of hearing protectors in noisy environments, thus minimizing the risk of impaired hearing amongst many other negative side effects.
However, with new equipment comes new risks. Hearing protectors does not only attenuate unwanted sounds, it also lowers the perceived loudness of warning signals, co-workers’ voices and other signals one might want or need in order to perform their work safely/effectively. The protectors also affect the users’ unconscious behaviour, as indicated by some researchers, such as Zhong & Yost (2013) and Lane
& Tranel (1971).
This master thesis project aims to investigate how hearing protectors affects the user’s ability to identify and recognize sound. The project is executed in cooperation with 3M Peltor and LTU in a mission to create a new method for measuring the performance of level dependant hearing protectors.
1.1 BACKGROUND
For certification of both passive and active hearing protectors there are well- established measurement methods which are recommended by the Swedish Work Environment Authority for properties important to avoiding hearing damage (Swedish Work Environment Authority, 2015). These measurement methods are insufficient for assessing the situational awareness of users of level-dependent hearing protection since passive protection constantly attenuates the sound level, while level dependent amplifies the signal depending on its strength. The aim of this thesis is to contribute with information to support the development of a new measurement method for sound identification with level dependent hearing protectors. The focus is on sound identification in the environment of processing industry. This investigation is based on literature surveys, interviews, and polls. Experienced HP- users were the subject for the project.
1.2 STAKEHOLDERS
This thesis is a request from 3M Peltor and LTU, the two main stakeholders developing a new method for measuring hearing protectors. Information regarding the effects of safety equipment in general is useful to the developers of said equipment, but also authorities in charge of occupational safety.
Parallel to this project, another master thesis is done for 3M Peltor and LTU by a student, also from Industrial Design Engineering. In that case focus is on the development of a test that will measure the sound identification ability when using level dependent hearing protectors.
1.3 OBJECTIVE AND AIMS
The purpose of this thesis is to contribute with information regarding the ability to identify sound whilst wearing hearing protectors. The strategy is to:
1. Study the current research regarding the effects on sound identification from hearing protectors.
2. Conduct interviews based on the literature to see if the experience from using hearing protectors in Swedish process industry corresponds with what the research indicates.
3. Sample the sound environment from a number of Swedish process industries to identify scenarios where level dependant hearing protectors are useful or problematic.
The needs from different hearing protection users will be identified and information with respect to the design and development of level dependent hearing protection will be collected. The project is being implemented to ensure the development of safer and better hearing protection. The goal is to find the answers to the following research questions:
• How does the ability to identify sound differ between passive and active hearing protectors?
• What type of signals is the user exposed to?
• How can the unconscious behaviour of the HP-user be studied?
• How do the users of hearing protectors perceive the relation between hearing protector usage and accidents or incidents?
• What hearing protector-related needs do the users have?
The results should contribute to the development of a test to assess the sound identification capability of the hearing protectors.
1.4 PROJECT SCOPE
This project was conducted over the spring term of 2020 and is about the effects on sound identification whilst wearing hearing protectors. It focuses on the risk associated with the phenomenon, specifically in Swedish process industry. Sound identification should not be confused with the wider term situational awareness which involves all our senses, hearing being one of those. Other properties such as attenuation and sound localization are not covered in this thesis.
This is a thesis within the field of occupational safety in the work environment, focusing on sound identification when using level dependent hearing protectors.
Therefore, parameters regarding the comfort and design of hearing protectors will in a large extent be ignored, if it is not related to user safety.
1.5 THESIS OUTLINE
Chapter 1, Introduction, presents the thesis’ background and purpose, together with the objective and aims of the project.
Chapter 2, Context, presents the company 3M Peltor and gives a brief introduction on how HPs and the LDF works.
Chapter 3, Theory, presents the theoretical framework used in the project.
Chapter 4, Methods, describes the methods and how they were used in the project.
Chapter 5, Results, presents the results from the interviews and polls.
2 Context
This chapter describes the current state of the art regarding hearing protectors, how they are used, and the measured effects they have on workers’ safety.
2.1 CURRENT STATE
The following is a description of the current work done in the field of hearing protection and occupational safety and information retrieval.
2.1.1 3M Peltor
3M Peltor is constantly working on improving HPs and to ensure the safest possible product for the user. 3M Peltor states three abilities humans have that is important to preserve in HPs.
1. Sound detection
The ability to detect whether a sound is present or not.
2. Sound identification
The ability to correctly identify what type of sound is present.
3. Sound localization
The ability to correctly localize the origin of the sound.
One might think that the better protector is the one that attenuates more than the other, but sound attenuation is only one of the parameters constituting a good hearing protector. 3M Peltor strives to not only attenuate loud noise, but actually reflect the desired sound to the user, thus increasing their situational awareness.
They do this with level dependant hearing protectors, that uses microphones and speakers to amplify lower sound levels while still attenuating noise to a safe level.
The list of demands that would constitute the “best” possible hearing protector is still developing, since the characteristics of good sound and safety-needs differ between users. It is, however, known that the ability to identify sound is compromised when occluding the ear, and it is believed that active hearing protectors to some extent can restore this ability whilst protecting the user.
2.1.2 The Swedish Work Environment Authority
The Swedish Work Environment Authority sets up regulations regarding workers’
safety. One of these regulations involves the usage of hearing protectors. In the brochure, Buller och Hörselskydd, ADI 344 (2015), the rules regarding noise in the workplace are summarized. The employer is obliged to ensure that the workers use hearing protectors when the average sound level in an 8-hour workday; exceeds 85 dB(A), if the loudest sound exceeds 115 dB(A), or if there exist short impulse sounds with a sound level exceeding 135 dB(C).
If the average sound levels in the workplace is higher than 80 dB(A) and shorter sounds peaks at 135 dB(C) or louder, the employer is obligated to inform the employee about the risks associated with noise and to offer hearing protectors.
These types of rules and regulations seem to work, since the percentage of work- related problems associated with hearing loss or hearing damage is low relative to other types of problems, according to The Swedish Work Environment Authoritys’
brochure Work-related Issues (2018) as can be seen in Figure 1. Noteworthy is that there are jobs where hearing protectors would be desired by the employee, but not suitable for the task at hand, such as kinder garden teacher, which might explain some cases of hearing damage.
Figure 1. Work-related issues, 2018
One problem that seems to arise from the usage of passive hearing protectors is the phenomenon of temporarily removing the protectors in order to better communicate with co-workers. The removal of hearing protectors with 30 dB attenuation for only 15 minutes during an 8-hour workday would be equivalent to those protectors only having a 15 dB attenuation. This is a problem since the choice of hearing protectors is based on the sound level in the environment, so using hearing protectors with 15 dB attenuation where 30 dB is needed would quickly increase the risk of hearing damage. This is a phenomenon that level dependant hearing protectors could help to solve, since the microphones and speakers are there to assist the users’ communicative needs.
Figure 3. Linear attenuation effect from a passive HP with 20 dB attenuation.
2.2 LEVEL-DEPENDENT HEARING PROTECTORS
This chapter gives a brief explanation to passive and active hearing protectors work.
Passive hearing protector muffs (Figure 2) attenuate the SPL by absorbing the energy in the plastic shell and letting the sponge part of the product act as a seal and spring on which to vibrate (verbal communication with 3M Peltor). The
attenuation is linear, as shown by Figure 3.
Actual SPL
Perceived SPL
Figure 2. 3M Peltors' passive HP, X4A.
Figure 5.Illustration of how the amplification varies with respect to the ambient noise SPLs in LDHPs. The orange line represent passive HPs attenuation and the black line represent the sound detected and played back in the speakers by LDHPs.
LDHPs (Figure 4), attenuates the sound pressure level in the same way, but also register lower SPLs via microphones and reflects them back to the user.
The hearing experience is therefore not linear like passive HPs and is illustrated in Figure 5.
Figure 4.3M™ PELTOR™ ProTac™ III Slim Headset
3 Theoretical framework
In this chapter, the theoretical framework for the project is presented, starting with relevant theories from the field of industrial design engineering relevant for occupational and safety and specific theories needed for this project.
3.1 INDUSTRIAL DESIGN ENGINEERING
What is Industrial Design Engineering? There really is no one absolute description of the field, and my anecdotal experience is that the definition depends on who is asked. Wikipedia defines industrial design as a type of design process specifically applied to products meant to be produced in larger quantities, referring to Heskett (1980). While not untrue, this might be an insufficient definition of the field. Other literature in the field seem to have a large focus on the people involved in the technology being developed. Some titles from the field are “Needfinding: The Why and How of Uncovering People’s Needs” by Patnaik & Becker (1999), “Arbete och teknik på människans vilkor” by Bohgard et al. (2015) and “Designers and Users:
Two Perspectives on Emotion and Design” by Norman (2003). The focus on the human is undeniable while not excluding the relevance of the process parameter.
The human presence in the industry is something that needs to be considered when designing both products and the production system. Products can be more or less user friendly and working in the industry can be more or less rewarding, damaging, or stressful. The industry can therefore be designed with the human as a frame of reference. One step in the right direction would then be to define Industrial Design Engineering as the process of designing a product, service, or a production system with regards to the human needs in every step of the way.
So how is this thesis connected to industrial design engineering? The process industry is, when not tended to properly, a place of risk. Despite dedicated recourses to ensure a safe work environment, people still go on sick leave and in some cases die as a result or their work. Hearing protectors of all sorts have the potential to reduce the risk of hearing impairment and is therefore a relevant subject matter of industrial design.
3.2 THE AUDITORY SYSTEM
In order to analyse the risks associated with an occluded ear, we must first understand on a basic level how the ear works. This is important in order to conclude whether a certain kind of hearing protector is useful or not.
What we perceive as sound traveling through a medium, is really only pressure differences in that medium, first registered by the eardrum. The eardrum transforms the pressure differences into kinetic energy that activates malleus, incus, and stapes (Silbernagl, 1991). These bones amplify the energy into the cochlea where the kinetic energy is transformed into electric signals, registered by the brain.
The auditory system also has the ability to determine a sound source’s location. This is called sound localization and is, according to H.P. Wallin et.al (2014) mostly the result of sound level and time differences between the two ears.
The stapedius muscle in the middle ear tightens when the ear is exposed to high
amplitude, low frequency sounds, protecting the cochlea to some degree in loud environments. However, this reflex is not fast enough to attenuate impulse sounds, and the muscles grow weary quickly in high frequencies. This is important to know when researching and using hearing protectors.
3.3 OCCLUSION OF AUDITORY SYSTEM
Hearing protectors are commonly used in noisy environments to lower the risk of hearing loss. However, new problems arise when the hearing protectors attenuates sound levels linearly, meaning that all the signals in the workplace are less noticeable, even warning signals and regular communication between co-workers.
A Bockstael et. al (2009) show that the HPs performances is largely based on what type of background noise is present. The article describes the speech recognition ability in different sound environments while using different types of hearing protectors. The experiments performed indicated that passive protectors can only help the communication in very noisy environments (>85 dB(A)) more than the unoccluded ear or with active hearing protectors. It is important, however, to notice that this is the result of one study from 2009 with hearing protectors not necessarily from 3M Peltor (manufacturer not acknowledged in article). In a sound environment with fork-lift trucks, an open ear is preferred when it comes to speech recognition, but a level dependant hearing protector with low output is the best occluded option. In the experiments conducted in the article, the output from the speakers is clearly a relevant parameter for speech recognition. Minimal amplification seems to be better than maximal amplification in four out of the five different sound environments tested. According to the authors, this is probably due to excessive amplification, leading to a distortion of the relevant signals. This phenomenon is also found in a study by Azman & Yantek (2010), and a lower degree of amplification is also what is recommended by 3M Peltor (from personal communication).
Speech intelligibility clearly deteriorates with an occluded ear in noisy conditions, but level dependency seems to be able to improve communication in this type of workplace.
There are several examples of how hearing protectors in different ways negatively affect the user. Brown, Beemer, Greene, Argo, Meegan & Tollin (2015) tested four HPs, two passive and two active ones, all of which performed worse than an unoccluded ear when tested for sound localization, speech recognition and tone detection threshold. The one exception being one of the active HPs which increased the tone detection threshold and maintained the quality of speech recognition. Kårekull & Johansson (2019) show the difference in REAT between the same HP with the LDF off vs. on. Figure 7 shows this for 3M Peltors’ HP, 3M™ PELTOR™ WS Alert XPI. It is clear from Figure 7 that this HP is beneficial to the user regarding sound detection.
The deviation around 103 Hz with LDF ON means that sounds with this frequency is more attenuated than other frequencies. This is, according to Wallin et. al (2014), about the same scope of frequency where humans speak (Figure 7). This might suggest that some words would be difficult to distinguish when communicating with this HP.
Figure 6. REAT ON/OFF for 3M™ 3M PELTOR™ WS Alert XPI
Figure 7. The Speech spectrums main area. A = Scope of fundamental tones. B =Vowels. C = Voicing consonants (e.g. m, n and v). D = Voiceless consonants (e.g. s, t and k)
3.3.1 Sound localization
Sound localization is covered briefly in this thesis, since the interviews conducted did include a part where the subject was asked to discuss the localization ability. In a study conducted by R. Mlyński and E. Kozłowski (2016), sound localization is tested with 3M Peltors’ LiteCom 111 level dependent earmuffs. The signal to localize was an overhead crane. The unoccluded ear was able to correctly localize ca 75% of the unmasked sounds. This decreased to ca 66% when the signal was masked with ambient noise. When using 3M Peltors LiteCom 111 without the level dependent function (passive mode), the results dropped to 44% and increased to 57
% when using the level dependent function.
3.4 ACCIDENTS AND INCONVENIENCE IN THE WORKPLACE
In order to understand what role LDHPs can have in Swedish safety work, we must understand the problems at hand in the Swedish work environment. Are there currently accidents occurring that could have been prevented or hindered with LDHPs?
A rapport done by the Swedish Work Environment Authority, Work-related issues (Arbetsorsakade besvär, 2018), shows that 3 out of 10 workers in Sweden suffers from some sort of work-related issue. 65 % of the people with work-related issues from something other than accidents, claimed that high workload is the root of their issues.
Accidents:
The most common type of accident in Sweden 2018 was “person falling”, meaning that a person at work has fell, slipped, or miss stepped and as a result from falling has been injured. This is the cause of ca 30 % of workers accidents. The second most common cause for accident is “overload”, meaning a failed heavy lift or a torn ankle. Hearing loss or anything hearing related is not listed in Figure 6. This does not mean that the usage of HPs was not a contributing factor to the accident. Loss of situational awareness is a hypothesis formulated by both 3M Peltor and LTU to be a danger in some work environments. However, no support for the hypothesis has been found during the literature survey.
Figure 8. Accidents in the workplace, 2018.
From talking and discussing potential benefits of wearing LDHPs instead of passive HPs with both 3M Peltor and LTU, the scenario of not hearing a truck is used as an example. It is hypothesized that one might not hear a vehicle, e.g. a forklift truck when visual information is limited, leading to an accident. No exact statistics of the causality of the truck accidents are published on the Swedish Work Environment
(https://www.av.se/produktion-industri-och-logistik/fordon/truckar/). Falling goods and spraining ankles is also not unusual. On the webpage where the Swedish Work Environment Authority is talking about the potential dangers of moving objects, the usage of HPs are not mentioned, nor the scenario where a person is hit by a truck.
Sick leave and inconvenience:
As shown by Table 1, 28 % of workers in the process industry have in the last 12 months gone on sick leave as a result from work related issues. This is mostly caused by accidents and is covered in previous chapter.
Table 1. Percentage of employees with work related issues leading to sick leave in the latest 12 months, sorted by field of work and gender, 2018. (Arbetsorsakade besvär, 2018)
The Swedish Work Environment Authority also documents general inconveniences caused by work. This is important when talking about the effects of hearing protectors on the user. According to the report from 2018, 38 % of Swedish workers in processing industry (and transporting workers) reported having sleep disturbances cause by work. This seems high, and might be, but compared to other fields of work it is among the groups who sleeps the best. Managers and jobs demanding a university degree are the groups that rapport the highest sleep disturbance numbers. 69 % of managers and 61 % of university graduates rapport having problems sleeping. This does not mean that there is no problem in the process industry. According to the Swedish Work Environment Authority, a noisy workplace can be a stress factor even when the noise in not very loud, leading to sleeping problems. A bigger problem in the processing industry is tiredness. 64 % of
processing workers rapport tiredness as an issue from work. It makes sense that a person gets tired from working, but when there is a clear correlation between the symptom and the source of the problem, there might be cause for action. 39 % rapports having headaches from working in the processing industry.
When exposed to low frequencies, the risk for headaches goes up, according to the Public Health Authority (https://www.folkhalsomyndigheten.se/publicerat- material/publikationsarkiv/o/om-ljud-och-buller-/?pub=60517). This is cause for suspecting low frequencies being described in the interviews.
18 % of process workers reported having problems with hearing, such as tinnitus, hard of hearing or hyperacusis, a testament to the efficiency of hearing protectors.
The most reported problem was physical pain, which 76 % of process workers had.
3.5 WORK ENVIRONMENT
It is important to tend to the human needs of a healthy work environment, not only to prevent quantitative accidents or injuries. Bohgard et al. (2015) states that people who have been working in an unhealthy environment for a significant time, usually develop adverse behavioural patterns such as anger and recalcitrance. Employers must therefore be aware of the needs of the employees in order to contribute with a healthy work environment. So how can hearing protectors be a part of a better work environment, more than just protecting the hearing? That would depend on the needs of the user. Several papers, one being Azman & Tantek (2010), rapport on the problem of not being able to communicate with colleagues in the noisy work environment due to the hearing protectors.
3.6 DATA COLLECTION
According to Osvalder, Rose & Karlsson (2015), data collection methods are useful for extracting data in a technical system in which humans are involved. Gathering data from these types of systems can be done in multiple ways e.g. interviews, observations, and polls. Since this thesis is a method study, this chapter briefly presents theories regarding how these methods work and are supposed to be used.
How they were used in the project is presented in Chapter 4 and discussed in Chapter 6.
3.6.1 Interviews
According to Osvalder et. al (2015), interviews are the most fundamental of all methods for gathering information about what people think and feel about a certain subject. Interviews are recommended for collecting qualitative information, rather than quantitative. Knowing this is key for planning the list of questions that are to be asked. Osvalder et. al (2015) also states that it is important to clarify the purpose of the interview, because the purpose decides what type of interview is to be conducted. Osvalder et. al (2015) presents three types of interviews: structured, unstructured, and semi-structured.
The structured is useful if quantitative type information is needed, e.g. how many years someone has worked at the company or how satisfied they are with their boss, on a scale from one to five.
If the interviewer seeks to understand the opinions of the subject, an unstructured
been foreseen. In this project, a question in an unstructured interview might be:
“Do you feel affected by your hearing protectors?”
and in a structured interview, that question might instead be formulated as a statement such as:
“In what way is your ability to identify sound affected whilst wearing your hearing protectors?”
1. 2. 3. 4. 5.
Impaired Unaffected Reinforced
3.6.2 Polls
Polls are described by Osvalder et. Al (2015) as a structured interview where the interviewer is not present. The three main areas of usage are the following:
• Collecting data from a large fleet of people
• Collecting data from people who are difficult to get contact with
• Validating results from previous interviews
The subject is usually given a scale or preformulated answers to choose from, leaving no room for development. Some questions can give the subject the possibility to freely formulate an answer, though this should not be overused since it makes the questionnaire more challenging.
3.7 UNCONSCIOUS EFFECTS FROM SOUND ENVIRONMENTS
There is some research that indicate an unconscious change in behavior, depending on the present sound environment. Awareness of these effects are important when analyzing interviews and polls since it might give an additional perspective on some answers from the users.
Zhong & Yost (2013) showed that there is a connection between auditory queues and postural stability. In other word, the presence of sound can be beneficial to the user when it comes to posture and balance. This is not a heavily investigated subject and the authors clarifies that more research is necessary to understand exactly how humans’ balance and posture is affected by auditory queues. The article reported a 9
% decrease in mean head sway when a loudspeaker played noise in front of the subject (eyes closed). Another example of how humans are unconsciously affected by sound is the Lombard effect, the tendency to change once voice level depending on the sound environment, as discovered by Lombard E. in 1911 (Lane & Tranel, 1971).
There are several indications that humans’ cognitive capabilities deteriorate in a noisy environment. Marsh, Ljung, Nöstl, Threadgold and Campbell (2015) present results indicating this phenomenon by having subjects listening to a list of words that were to be recalled afterwards, once without noise and once with noise. A
decreased ability to memorize the list in a noisy environment was found and is in line with previous research indicating that it is more difficult to encode and process spoken information in a noisy environment.
4 Method and Implementation
This chapter describes the overall process of the project and how the methods were used.
4.1 PROCESS
The project started with getting familiar with the areas the thesis covers. After a Skype meeting with 3M Peltor where the project was discussed from their point of view, the literature study could start. This was done by searching for papers related to the subjects of hearing protectors, LDF, sound recognition etc. After ca. two weeks into the project a small presentation was held for 3M Peltor via Skype, where I briefed them on the researched topics. A process for the project was chosen at this point. The process must be compatible with the task at hand. This project required me to conduct interviews based on theories not necessarily addressed in the education of Industrial design engineering. It seems reasonable to choose a process that plans for change when new knowledge is attained. The project spiral, as described by Ranhagen (1995) is a process that does this.
4.2 PROJECT PLANNING
The planning was done with the project spiral in mind and there was plenty of room for searching more papers and re-planning the project. The plan was visualized in a Gantt-schedule with three main project phases, Figure 9. One cell is one week, and the three main phases are visualized in the first column in three different shades of green. The sub-phases are intentionally overlapping each other where some iteration is suspected to occur. One example is how the literature study stretches all the way to the data collection phases. This is due to the suspicion that new knowledge would be attained from the interviewed subjects that would require more articles and references.
Due to a pandemic outbreak of the virus Covid-19, visits to companies was not an
option. A larger focus was instead put on interviews and theory to ensure a scientific weight to the thesis. The interviews were conducted via telephone.
4.3 METHODS
The following subheadings are descriptions of the methods used for deeper understanding of how hearing protectors are perceived by the users.
4.3.1 Literature Review
In the first phase of the project spiral, academic databases were used in order to gather as much research as possible to read. This was done, not to find exact references, but to familiarize myself with how far the research has come regarding HPs, sound recognition, situational awareness, and work safety. Some papers written by 3M Peltor was e-mailed from their representatives.
The main search engine used was Scopus. On occasion, Google and ResearchGate was used mainly to find open access versions of articles found in the main search engines. The key words used to find articles, and the result numbers are presented in the results chapter.
The listed objectives with the project acted as inspiration for searching for literature and was continuously revised. This means that the search and usage of the articles was continuously changing throughout the project.
4.3.2 Interviews
Since the project to a large extent is about the perceived experience of hearing protectors it would be logical to choose unstructured interviews as a method of choice. However, some quantitative measurements might still be needed and so the third option, semi-structured interview, is reasonable to consider. This allows the interviewer to ask questions based on answer in order to explore a topic. Better yet might be to conduct an unstructured interview with a separate poll that the subject fills out parallel to the interview. This would allow for a bigger data collection only if a company would be willing and/or able to send out the questionnaire to a large number of employees, which none was. The chosen strategy was therefore:
• Conduct semi-structured interviews via phone with a goal of ca 20 subject.
• Ask a mix of qualitative and quantitative questions and let the subject discuss the answers.
Since the interviews was done by phone, taking notes on a computer was not distracting for the subject, so no recording was made.
• Send out polls to companies’ head safety officers and ask them to send the poll to as many regular headphone users as possible.
• Leave some room for free answers in the polls to see if any new narratives are exposed. If a significant amount of people fills out the questionnaire, statistical methods can be used in order to show how likely it is that a point of view is held by a certain share of users.
4.3.3 Constructing the interview
According to Osvalder et.al (2015), it is advised to start of the interview with some easy questions that would not be a challenge for the subject to answer. Some of the questions asked in these interviews were:
“What kind of HPs do you use?”
“Are your HPs ever annoying?”
It is essential to know what to look for when conducting an interview. In order to know what questions had to be asked, I went back to the main objective of the project and looked at the questions that were supposed to be answered in the end of the thesis.
The first question
• How does the ability to identify sound differ between passive and active hearing protectors?
is relevant for the developers of HPs and for people working with work safety- questions. It is difficult to investigate this with only interviews, since people in general do not really pay attention to their sound identification ability, unless something dramatic happens as a result from being cheated by the HPs. To deal with this, the first question and the second question
• What type of signals is the user exposed to?
were strategically asked together, with a short introductory line for contextualisation:
Interviewer: Signals are used in order to inform people about something in a specific area. What type of signals exist in your work environment?
Subject: Answers…
Interviewer: Could you imagine that these signals could be mistaken for each other or something else?
This was supposed to get the subject to really think about their experience of using hearing protectors and how they are affected by it.
The third question that needed to be answered in the thesis
• How do the users of hearing protectors perceive the relation between hearing protector usage and accidents or incidents?
was handled in the interviews by breaking it into three questions. The first one being almost a silly question, but nevertheless important:
“Do you feel safer or less safe whilst wearing HPs? Why?”
Followed by:
“Do you feel that there might be risks added as a result of wearing HPs?”
Followed by:
“If not risks, what other possible problems do you think can arise from wearing HPs?”
The subject answered between every question. The fourth question
• What hearing protector-related needs do the users have?
might seem as a trivial one, but it is impossible to design a good product if the needs of the customer are not investigated. If asked plainly, “what do you need from your HPs”, the answer would most likely be “protection for my ears” in almost all cases.
Therefore, the interview ended with the two questions:
“What is, in your opinion, the most important feature in an HP?”
and “What is, in your opinion, the second most important feature in an HP?”
These questions allow the subject to freely express what is most important to them.
Looking at the scope of the project, it is however reasonable to create a fixed set of features for the subject to rank. This would ensure that the users’ needs are addressed, while still forcing a priority list useful for 3M Peltor. The features that would be interesting to have the subject rank is, sound identification, speech recognition and sound localization. Speech recognition could be viewed as a subcategory of sound identification, so it would be logical to see this as one feature.
In this context, however, the two should be separated since there is a big user experience difference in correctly identifying danger and being able to effectively communicate with colleagues.
In the interviews, a simple explanation of the three basic properties of HPs was presented, after which the subjects were asked to grade them on a scale from 1 to 10, one being not important at all and ten being especially important. If the subject gave the same grade to two or all features, they were asked to rank them to each other, forcing a value hierarchy.
Since no visits to the workplaces could be done, the subjects were asked to describe the sound environment in their workplace. At first, they were asked to freely describe the sound environment. After this, more specific questions were asked, such as
“What sound has to be attenuated in your workplace?”
“What sounds do you need to hear?”
The earlier question about what type of signals the user is exposed to was also useful for a general description of the sound environment.
4.3.4 Constructing the questionnaire
The questionnaires’ purpose was to complement and validate the stories told from the interviews. In an early interview, one subject said that they almost never activate the LDF in the HPs which lead to one question in the questionnaire being:
“Is your LDF mainly on or off?”
The intention of this question was to compare users’ habits to their thoughts on sound identification and their overall attitude towards HPs.
The subjects were also asked about their situational awareness. This was done by asking them to grade their abilities on a scale from 1-an odd number. This is a common method for measuring peoples’ attitude and opinions on a topic and is called the Likert scale. An odd number of choices is commonly used to leave a middle option that suggest some sort of indifference on the topic. The Likert scale was used when the subjects were asked on their abilities to identify origin of sound, identifying sound and communicating with colleagues. The choices regarding communication with colleagues differed slightly from the two about sound identification and localization.
4.4 IMPLEMENTATION
This chapter describes the methods used for analyzing the data gathered from interviews and polls. The data is collected to get a perspective from the HP users’, and the methods used should reflect the purpose of the study. A method for analyzing interview transcripts was found and used. The polls were sent via e-mail to principal safety representatives who in turn sent it to their HP-users.
4.4.1 Data analysis
Data collected from interviews and polls, in the field of Industrial design engineering tend to differ from data collected from the field of Computer science.
To some extent, numbers and other quantitative metrics are attained from interviews, but it is not comparable to data from e.g. a GPS-satellite. In an interview with Ann Blandford (https://www.interaction- design.org/literature/article/how-to-do-a-thematic-analysis-of-user-interviews) she says that the following six steps are useful in analysing data from interviews.
1. Familiarization
This step is about reviewing the transcription from the interviews and to familiarize oneself with the data. Blandford argues that the transcription is within itself a way of getting familiar with the data. This is a good time to make notes on topics that is interesting for further investigation. In this project, a method for familiarization was to transcribe the notes taken during the interviews.
2. Assign descriptive codes to the data
Make more structured notes in the transcription that describe what has been said.
These single bits of information are called codes and are a way of creating an overhead view of what is being said.
Figure 10. The process of coding the transcript
3. Search for patterns
Sort the single units of code into themes. It is expected to iterate over this step, sorting the codes in different ways in order to arrange them in a way that makes sense. This is a more descriptive step than step 2 where bits of seemingly useful information are isolated.
4. Review the themes
In step 4 the themes are more deeply analysed. Are there contradictions in the themes, or do they overlap? Compare the codes to the themes and see if the categorizations still make sense. Iterate this process over the different sorting categories that has been made in previous steps and stop when nothing relevant is added anymore.
5. Define the themes
The themes must be defined and named. This is about finding the essence of the theme and is supposed to tell a story about why the theme is interesting. By now, the analysist should be able to contextualise the themes and to speak clearly about each themes’ relevance.
6. Producing the report
Step six consist of writing the report and gives some advice on how to structure the paper. For this thesis, I present my work in a structure I find reasonable for the project, so this step is overlooked.
This method of analysing data from interviews is gathered from the Interaction design-foundations’ website, so it stands to reason that the method should not be viewed as a scientifically proven method for a successful analysis. However, it is described by Ann Blandford, professor of Human Computer Interaction at University College London, which gives the method some scientific weight.
It makes sense to use this method as a base for the analysis and to complement it by looking for pieces of the interviews that are of interest in regards to the initial research question, posed in chapter 1.3 Objective and Aims.
4.4.2 Collecting quantitative data
The polls were sent via e-mail as an anonymous Google Questionnaire. The data is visualized automatically with the Google tool and was also exported to a Google sheet for the possibility to further analyze the data. The answers were compared in ways dependent on what is searched for at the moment. In the Google Sheet, the answer columns could be sorted and placed next to other answer columns in order to search for patterns in different HP users.
4.5 METHOD DISCUSSION
In this chapter, the methods relevance and implementation are discussed. The purpose is to validate the scientific work that has been done in the project.
4.5.1 Interviews
The more subjects interviewed the better, to a certain degree. I would argue that the information gathered from the interviews is sufficient for getting an idea of what is needed from the HPs. The interview is also scalable, meaning that 3M Peltor is free to continue interviewing users, thus gathering more personal experiences from HP usage.
4.5.2 Polls
Just like the interviews, the questionnaire is scalable and free for any HP developer to use. This serves as a ground for a large collection of data. Efforts were made in this project to send the questionnaire to large number of users. A problem that arose was that the companies with many employees using HPs do not assign operators with work emails, limiting the possibility to send the poll to hundreds of HP-users.
This would open the possibilities to draw more scientifically based conclusions based on statistics.
5 Results
5.1 RESULTS OF LITERATURE STUDY
The results of the literature study are partly presented in chapter 3. Theoretical framework. This chapter summarizes the information gathered from the literature study and how it was implemented in the project. The research questions initially posted in the thesis served as a base for the literature search. The discovered literature then served as a medium for translating the research question into questions that would be interesting to ask interview subjects and to measure with polls. The results from the Scopus searches are presented in Table 2.
Table 2. Results from Scopus search
When looking at the statistics, it is difficult to argue that passive HPs are a clear contributing factor to accidents and incidents in the Swedish processing industry.
Many speculations regarding the negative effects of passive HPs are found, both from discussing with 3M Peltor and LTU but also from reading literature from the Swedish Work Environment Authority. The scope of a thesis is to small to determine whether HPs are a statistically significant contributor to workplace accidents. It could be determined if the Swedish Work Environment Authority chose to collect the necessary data, but to many details are needed (exactly what happened, why, and what equipment was used). In this thesis, the users’ perspective was explored with the scientific literature as a starting point. Does the research match the experience?
From reading the current literature, it clearly seems like the usage of HPs passive or active, deteriorate the situational awareness of the user. From the statistics gathered by the Swedish Work Environment Authority, there’s not much that can be concluded regarding workers’ safety affected by HP usage. Working in noisy environments can lead to headache and stress, which is clearly present in the processing industry, but not as much as physical pains. The question remains whether the usage of HPs is a contributing factor leading to accidents. This is not investigated by the Swedish Work Environment Authority but is investigated via interviews and polls in this thesis.
5.2 RESULTS OF CONTEXT IMMERSION
These are the results of the interviews and polls. First, the results from the analysis method are presented. These are the results regarding what codes and themes were found by using the method for data analysis and are presented in the following subchapter 5.2.1 Data analysis. The findings from the interviews are presented in the next subchapter, 5.2.2 Interviews.
Key words Results Open access
Level dependent hearing protector 54 10
Sound identification hearing protectors 11 1
Sound recognition hearing protectors 24 1
5.2.1 Data analysis
The method used for analysing the interviews resulted in several codes, sorted in themes. Each theme is assumed to tell something of importance, which may or may not be the case. This is discussed further in the discussion chapter, where the statements from the interviews are compared with findings from the literature study.
Codes
81 codes were created from the transcripts of the five interviews. Some examples of codes are:
• HP warm in summer
• Trouble communicating
• Machines create the noise
• Need to hear in order to do the job
• Not bothered by HPs
These codes were read through several times and sorted in themes that would act as topics covering all the codes.
Themes
The 81 codes were sorted in seven themes. An eighth theme was initially found but finally removed. The them was Communication and was removed since all the codes regarding communication was related to electronic listening, an existing theme. The themes were later renamed to titles more pointing to the purpose of the thesis. These titles are:
• Issues
• Sound identification
• Sound environment
• Unwanted sounds
• Feedback sounds
• Electronic listening (LDF)
• Risk behaviour
The interviews were further analysed with these themes as a foundation and a starting point.
5.2.2 Interviews
Five interviews were conducted with HP-users from a processing company in Linköping, Sweden. All subjects used LDHPs from 3M Peltor regularly, though one mainly used casted hearing plugs (passive in ear protection). The interviews lasted about 20 minutes each after the subject was informed that they were free to leave at any time and that their names would not be disclosed in the rapport or presentation.
Findings from Data Analysis
These are the results from analyzing the interviews based on the data analysis described in chapter 4.5.1 Data Analysis.
Sound identification
Sound identification was introduced to the subjects as the ability to correctly determine what sound is present. Speech recognition and sound localization was also explained briefly. When asked to rank these three on their importance, one of the five said that sound identification was the most important feature of the three.
Four out of five said that speech recognition was the most important. This was further investigated in the questionnaire and is presented in the next chapter. The subjects were asked whether they think the signals at their workplace (trucks driving and honking, fire alarms, signals from the production line etc.) could ever be mistaken for another type of signal. Four said no and one said yes. The person saying yes had a roll within work safety and said that the machines have several signals that are used to communicate changes or that something is wrong. This means that the more machines, the more signals to keep track of. Whether this is a concern for HP developers is discussed in chapter 6.
Sound environment
One part of the interview had the subject describe their sound environment. Table 3 is a list of the adjectives used. Some are briefly explained by the author to contextualize what was being said. Some descriptive language can be lost in translation, so a table with the original adjectives used in Swedish is attached as appendix 2.
Table 3. Description of workers sound environment
Adjective Explanation
Machine noise
Industrial machine noise Powerful fans
Metallic sounds From steel drums used for a slicing process Rattling metal drums From same slicing machine
Howling machines From a vacuum packing machine
High sound peaks From products being thrown onto the production line
Intensely monotone
Hissing sound From packaging machine using gas to control the seal.
Unwanted sounds
The subjects were asked about the most problematic sounds that had to be attenuated in the workplace. Four out of five said that the general noise from the machines had to be attenuated, which is expected from the way the question was asked. When asked to develop their answers, different types of sounds were listed.
There seemed to be a frustration with the machines that does not create a monotone noise but rather more unpredictable sound peaks and rattling sounds, which three people pointed to.
One person raised concerns regarding the warning signals in the workplace, such as the signals informing about change in the production line. The same person did not raise any dissatisfaction regarding the LDF, which was regularly active. It stands to reason that this was probably brought up to say that those signals are too loud to listen to without HPs, not that higher frequencies are especially annoying with their LDHPs.
Feedback sounds
It is not unusual to use sound signals as a feedback in the workplace. Problems would arise if these relevant sounds were attenuated to the point where it is no longer heard. Sounds such as signals from the production line, fire alarms, and a hissing sound from a packaging machine were listed by the interviewed. An interesting perspective was that the regular noise from the machines was considered useful to know that the machine worked properly. Some said that the regular noise was needed, and some said that any deviating noise from the machines was needed.
Even when a machine is identified as defective, sound is relevant in the process of troubleshooting, e.g. to see if the problem is electrical or mechanical.
Electronic listening (LDF)
Some problems were highlighted with the LDF, the function supposed to enhance communication and situational awareness. All the interviewed used LDHPs but only three out of five had the LDF active most of the time. One chose to turn the function off, saying that high frequency sounds are too loud with the LDF on.
There was a general dissatisfaction with the LDF among the subjects. Three out of five reported three different issues with the LDF. The following are quotes from three different interviews.
“It [the LDF] enhances the rattling noise”
“High beeping sounds get through the LDF”
“The HPs can’t reproduce the sounds exactly. It feels different. Sometimes I remove the HPs to hear how it really sounds”
The two first quotes came from users who most of the time have the LDF deactivated.
Some problems with the amplification adjustment of the LDF and the radio were also highlighted. This was mainly an interface issue and was not further investigated.
Risk behaviour
It is well known that the processing industry is a place of risks, which is one of the reasons for the existence of the Swedish Work Environment Authority. Some behavioural patterns were discovered that matched what was said during the start up meetings with LTU and 3M Peltor.
Four out of five people said that they on occasion have removed their HPs in the noisy environment. When asked what the reason for the removal was, three subjects said it was to enhance the communication. One said it was to hear how the environment really sounds. This behaviour was also highlighted by one of the interviewed as a risk with wearing HPs in a noisy environment where
Another possible dangerous behaviour is the habit of listening to music with the HPs on but the LDF off. When the LDF is activated, the general sound environment is reflected to the user, creating a situation comparable to having a radio in a noisy room. When listening to music with the LDF of, the experience is not only similar to using passive HPs. The music creates a masking effect, making the background noise less noticeable. This was better investigated using polls and is presented in the next subchapter. In the interviews, however, some attitudes regarding safety was noticed. It was not unusual to refer to other people as risk factors, suggesting them to be less aware in the workplace than oneself. One quote from the question regarding potential risks with HPs is:
“I have worked here for so long that I’m aware of what is going on around me”
Issues
Some issues were presented when asked if the HPs are ever annoying or problematic. One issue brought up by several of the interviewed was the problem with glasses. A small gap is formed between the ear and the protective foam, lowering the attenuation effect. Having glasses does not, however, seem to influence the comfort of the HPs. One of the subjects working in more narrow spaces stated that the antenna sometimes got stuck on things which was a source of frustration to the subject. Several of the interviewed stated that the HPs are more uncomfortable during the summer when it is warm. Three out of five people brought this up as an issue without being questioned on comfortability or whether they are warm.
Results related to research questions
The plan was to complement the results from the data analysis by seeing what is left unanswered from the initial research questions. Results related to the following research questions are presented in this subchapter:
• How do the users of hearing protectors perceive the relation between hearing protector usage and accidents or incidents?
• What hearing protector-related needs do the users have?
Accidents and incidents related to HPs
There are some concerns regarding the usage of HPs. The idea that one might not hear a truck due to HPs was discussed early in the project as a potential danger. This was highlighted in one of the interviews but slightly rejected as a non-issue, since trucks in the specific workplace only drive in specific areas that are improbable not to notice.
Another concern was the ability to listen to music while working. There is some belief that this could lead to a decrease in situational awareness.
These were the only thoughts related to accidents and incidents, but some risks were highlighted that are not related to an immediate danger. If the HPs are not completely satisfactory the user will more likely act in risky ways, such as removing HPs to talk or because they are warm.
Other than what is listed above, there were no concerns regarding change in safety from wearing HPs. Frustration was a more prevalent topic of discussion during the interviews. Only one person said that their HPs are never annoying or frustrating, but even this person was able to list some problems with the product.
