Assessment of annoyance and cognitive fatigue of open-plan office occupants subjected to noise from two different activities Lucas Lenne 1 INRS 1, rue du Morvan, CS 60027, F-54519 Vandoeuvre Cedex Patrick Chevret 2

INRS 1, rue du Morvan, CS 60027, F-54519 Vandoeuvre Cedex Etienne Parizet 3 Univ Lyon, INSA Lyon, LVA, 25 bis av. Jean Capelle, F-69621 Villeurbanne

ABSTRACT An open-plan office can be organised into zones, in which different activities can take place (focused individual work, collaborative work, phone communication...) The ISO 22955 (2021) standard recommends acoustic attenuation values between zones that varies according to the activity of the emitting zone (which influences the noise emitted by the occupants of this zone) and that of the receiving one (in which the necessary concentration may vary). The objective of this study was to evaluate the relevance of these target attenuation values. 32 participants who had to carry out a focused individual activity (writing a press review) were submitted to two types of noise environment (representing the noise emitted from a breakout area or that coming from an area where people mainly have telephone conversations), at four signal-to-noise ratios (in relation to the own noise of the reception area). The exposure time was approximately 35 min, and participants completed questionnaires on their perception of the noise environment, cognitive fatigue (MFI) and mental workload (ICA). The presentation will show the results of this experiment and the practical consequences for the target values of the ISO 22955 standard, in the case of these two disturbing noises.

1. INTRODUCTION

Service economy, which is defined as the growth in the share of services in a country's overall production, is a phenomenon that has been very present in France for several decades and is tending to become widespread in Europe and throughout the world. This phenomenon is reflected in a significant increase in the number of large open-plan offices, which have the merit of encouraging exchanges between departments and communication between company employees. Nevertheless, this

1 lucas.lenne@inrs.fr 2 patrick.chevret@inrs.fr 3 etienne.parizet@insa-lyon.fr

Jai. inter noise 21-24 AUGUST SCOTTISH EVENT CAMPUS ? O ? . GLASGOW

type of workspace layout has its drawbacks, notably the fact that it leads to noisy environments, in which conversations at workstations spread to the entire work area and which can ultimately prove to be deleterious to concentration, generate cognitive fatigue in the medium term and, in the long term, lead to problems of stress, sleep disorders etc. The evaluation of the quality of the sound environment in open-plan offices is therefore a major issue in terms of occupational health and is a priority issue from the point of view of prevention. The acoustic quality of a working environment is complex to determine because it needs to take into account the perception of employees, which depends on individual factors. Nevertheless, there are physical factors, linked in particular to the acoustic quality of the premises, and human factors which can be clearly identified. Their effects can be estimated using acoustic measurements or questionnaires. For this purpose, ISO 22955 was drafted and published in 2021. The quality of the open-plan office environment can be assessed using this standard. It proposes two approaches, depending on whether a single activity is carried out in the office or whether several activities coexist there. In the first case, the standard recommends the use of indicators to characterise the acoustic quality of the room - the measurement of these indicators being defined by ISO 3382-3 (2021) - and the average acoustic level during activity. It also provides values for these indicators according to the activity carried out. In the second case, it proposes a new indicator - the attenuation of the A-weighted level of speech, noted DA,S - to be used between the closest stations of two neighbouring activities that interfere in the same office. The standard also proposes target values for DA,S depending on the two activities (one being considered as disturbing the other one). The method for estimating the target values is based on the idea of liveliness within activities [1]. In this paper, we propose an alternative approach based on the laboratory assessment of annoyance, cognitive fatigue and workload during the performance of a press review task as a function of the signal-to- noise ratio between the two activities in the disturbed area.

2. EXPERIMENT

This experiment took place in the semi-anechoic room of LVA (in Lyon). Participants were asked to carry out an office task while being submitted to a noise during 35 minutes.

2.1. Participants Thirty-two people (18 women, 14 men), recruited by an external company, took part in the experiment. The average age was 41 years and ranged from 18 to 65 years. The participants had at least a bachelor's degree and declared that they had no hearing loss.

2.2. Sound stimuli

Sound stimuli were made of a mixture of two types of noise. One was constant in all conditions. It represented the noise from a quiet area, in which people do a personal focused work. This noise was made of a quiet ventilation hum, plus some sounds representing the use of computers keyboards and mouses. Two other noises were used. The first one represents sounds which can be heard from a breakout area, i.e. lively discussions. It was created from internet samples. The second one was recorded by INRS in a call center; it is made of “half-conversations” held in a more even tone. Each of these latter two noises was mixed with the “quiet area” noise, after being filtered in order to have a similar long-term average speech spectrum. Four different signal-to-noise ratios were used: -6, -2, +2 and +6 dB. The presentation level of the “quiet area” noise was fixed to 41 dB(A). As a consequence, the overall level of each mixture varied from 42 to 48 dB(A).

2.3. Procedure

The experiment was carried out in groups of 4 people, in the semi-anechoic chamber of the Acoustic Vibration Laboratory of INSA Lyon (figure 1). Each group came for two half-days (morning or afternoon to respect circadian rhythms). Four tables were set up around an omni-directional source on the floor.

Figure 1: Experimental set-up. Over the course of half a day, participants were exposed to one of two sound mixtures for 35 minutes for each signal-to-noise ratio.

For each condition, they were asked to write a press review based on newspaper extracts relating to a specific recent theme (a different theme for each condition). They were informed that they would be asked to fill a multiple choices questionnaire later, in order to keep them involved in the task. At the end of each exposure, they were also asked to answer several questionnaires:

• A french version of the Multidimensional Fatigue Inventory questionnaire [2,3]. • A questionnaire adapted from the ICA (Individual – Charge – Activity) [4,5]. • An evaluation of three characteristics of the noise they experienced during the last exposure (loud, tiring, annoying), thanks to three Likert scales (8 points).

2.4. Results

Results will be presented during the conference. Preliminary ones show that, as an example, the increase of mental fatigue with signal-to-noise ratio is greater for the “breakout noise” than for the “callcenter noise” (figure 2).

Figure 2: Relation between signal-to-noise ratio and perceived mental fatigue, for the two types of

noise (mean values in their 95% conf. int.)

Mental fatigue N © N © N N N o XN a N BR N @ N ND N EB cattcenter Breakout -6dB -2dB SNR +2 dB +6 dB

4. CONCLUSIONS

The difference between the two noises appeared after a short exposure time (35 mn). They can be expected to be much higher in the case of a full working day exposure. Thus, the approach proposed by ISO 22955 standard seems meaningful. But this study could be used to adjust the signal-to-noise ratio targets mentioned in that standard.

5. REFERENCES

1. Vellenga S, Bouwhuis, T. & Höngens, T. Proposed method for measuring “liveliness” in open-

plan offices. Proceedings of 24 th International Congress of Sound and Vibration. London, U.K. July 2017. 2. Smets, E., Garssen, B., Bonke B. & de Haes, J. The multidimensional Fatigue Inventory (MFI)

psychometric qualities of an instrument to assess fatigue. Journal of Psychosomatic Research , 39 , 315–325 (1995). 3. Gentile, S., Delarozière, J.C., Favre, F., Sambuc, R. & San Marco, J.L. Validation of the French

“multidimensional fatigue inventory” (MFI 20). European Journal of Cancer Care , 12 , 58–64 (2003). 4. Galy, E., Cariou, M. & Mélan, C. What is the relationship between mental workload factors and

cognitive load types? International Journal of Psychophysiology . 83 , 269–275 (2012). 5. Galy, E. Consideration of several mental workload categories: perspectives for elaboration of new

ergonomic recommendations concerning shiftwork. Theoretical Issues in Ergonomics Science . 19 , 483–497 (2017).