0LAUK0 2018Q1 Group 2 - SotA Literature Study

Introduction

This page contains the initial State-of-the-Art literature study performed by group 2 for the course Project Robots Everywhere (0LAUK0). The main research subject was broken down into several relevant topics. This literature study is divided into the following relevant topics:

• How long does it take for RSI to take effect? (new week 6)
• How big of a problem is RSI? (new week 2)
• RSI issues: upper extremity problems (new week 2)
• RSI prevention: how to properly setup a computer monitor (new week 2)
• RSI prevention: how to use breaks to combat RSI (new week 2)
• Sensor Technology
• User Interface
• Repetitive Strain Injury
• Dutch Arbo legislation
• Existing products
• Motor Control
• Relevant developments at Eindhoven University of Technology

Each topic has a summary and a list of references.

Summaries

How long does it take for RSI to take effect?

In general, working with a computer for more than 6 hours a day could cause RSI symptoms (to be exact, upper limb disorders). Here, men can sometimes work a bit longer with a computer without getting the symptoms while women can already get symptoms when working 4 hours a day with a computer. Also high stress enlarges the probability of getting RSI symptoms. The first signs may be soreness, tingling or discomfort in the neck, arms, wrists, fingers or shoulders. These symptoms can occur when doing a repetitive task. It can take a few hours to a couple of days for the symptoms to develop, depending on the physical condition of the user.

How big of a problem is RSI?

In the Netherlands:

According to the federation of Dutch Trade Unions (FNV) 11% of all sick leave in the Netherlands is caused by conditions that are part of RSI (Federatie Nederlandse Vakbeweging, n.d.). In order to assess the seriousness of RSI in the Netherlands, we turn to TNS NIPO. This company specializes in public surveys. In this section we will use the results from the ‘national survey on working conditions’, using the editions from 2008 and 2013. These surveys are filled in by Dutch employees. Luckily for us, they include questions on RSI prevention at the workplace, and whether the participant suffers from RSI or conditions that have a similar effect on their health (chronic pain in their arms and/or hands, chronic pain in their back and neck). Let us summarize their results. We include the percentages of certain business sectors as we believe that these sectors include desk work.

In 2013 a combined total of 17.2% of 22,610 participants across all sectors answered that there were no or insufficient measures against RSI at their workplace. We believe the following sectors include desk work and are therefore important indicators for the necessity of RSI prevention. The percentages listed below are the combined sums of participants answering that there were no or insufficient measures against RSI at their workplace in their respective business sector. All data originates from van Zwieten, de Vroome, Mol, Mars, Koppes & van den Bossche (2013, p. 73).

No or insufficient measures against RSI at my workplace (2013) (N = 22,610) (% answered yes)

• ‘Information and communication’: 28.1%
• ‘financial institutions’: 25%
• ‘business services’: 22.2%
• ‘public administration’: 22.9%
• ‘education’: 19.3%

Participants were also asked to indicate whether they suffered from one or more chronic impairments. From a sample of 22,901 participants, across all business sectors a combined total of 5.7% participants answered that they suffered from problems with their arms or hands (also caused by arthritis, rheumatism and RSI). From this same sample, across all business sectors a combined total of 10.3% answered that they suffered from problems with their back and neck (also caused by arthritis, rheumatism and RSI). The percentages listed below indicate how many participants answered that they suffered from one of the previously mentioned impairments. All data originates from Van Zwieten et al. (2013, p. 78).

Problems with arms or hands (2013) (N = 22,901) (% answered yes)

• ‘Information and communication’: 4.5%
• ‘financial institutions’: 4.5%
• ‘business services’: 5.8%
• ‘public administration’: 6.7%
• ‘education’: 5.9%

Problems with back and neck (2013) (N = 22,901) (% answered yes)

• ‘Information and communication’: 8.9%
• ‘financial institutions’: 8.9%
• ‘business services’: 10.6%
• ‘public administration’: 12%
• ‘education’: 10.7%

In 2008 a combined total of 19.4% of 21,469 participants across all sectors answered that there were no or insufficient measures against RSI at their workplace. The ‘Information and communication’ sector is missing from this iteration of the national survey on working conditions. The percentages listed below are the combined sums of participants answering that there were no or insufficient measures against RSI at their workplace in their respective business sector. All data originates from Koppes, de Vroome, Mol, Janssen & van den Bossche (2008, p. 70).

No or insufficient measures against RSI at my workplace (2008) (N = 21,469) (% answered yes)

• ‘financial services’ at a combined total of 23.2%
• ‘business services’ at a combined total of 25.7%
• ‘public administration’ at a combined total of 23.5%
• ‘education’ at a combined total of 19%

Participants were also asked to indicate whether they suffered from one or more chronic impairments. From a sample of 21,208 participants, across all business sectors a combined total of 5.5% of all participants answered that they suffered from problems with their arms or hands (also caused by arthritis, rheumatism and RSI). From this same sample, across all business sectors a combined total of 10.2% answered that they suffered from problems with their back and neck (also caused by arthritis, rheumatism and RSI). The percentages listed below indicate how many participants answered that they suffered from one of the previously mentioned impairments. All data originates from Koppes et al. (2008, p. 75).

Problems with arms or hands (2008) (N = 21,208) (% answered yes)

• ‘financial services’: 3.9%
• ‘business services’: 5.4%
• ‘public administration’: 5.2%
• ‘education’: 5.8%

Problems with back and neck (2008) (N = 21,208) (% answered yes)

• ‘financial services’: 8.6%
• ‘business services’: 9.5%
• ‘public administration’: 11%
• ‘education’: 10.7%

In Europe:

The European Agency for Safety and Health at Work includes the disorders that are part of RSI in their definition of Work-Related Neck and Upper Limb Disorders (WRULDS). In a factsheet published in 2007 the organization claimed that “almost two thirds of EU workers report being exposed to repetitive hand and arm movements … significant risk factors for work-related neck and upper limb disorders.” (European Agency for Safety and Health at Work, 2007). Although WRULD encompasses all work (such as construction work), office work aligns with a lot of the risk factors and dangerous activities mentioned in the fact sheet. These include working in awkward positions (which can be caused by using a misadjusted desk or uncomfortable keyboard), repetitive movements (typing), and prolonged work without opportunities for rest. The fact sheet mentions that improving the layout of the workplace is a prime candidate for preventing WRULDS. Our vision for the RSI robotic desk is in line with this advice.

RSI issues: upper extremity problems

RSI problems concerning lower arm, elbow, wrist and hand problems are quite known in the Netherlands. According to the Nationale Enquete Arbeidsomstandigheden (NEA) 2017 in the Netherlands nearly %37,3 of the population questioned (N≈40.000) have experienced some form of wrist/hand injuries and %30,7 arm/elbow injuries due to their work. Luckily for them the majority only experienced it once for a short amount of time, which leaves only %5,9 for wrist/hand injuries and %13,3 for arm/wrist injuries for a longer amount of time.

The NEA 2017 also specifies statistics concerning RSI-problems in the education industry:

NEA 2017: RSI-problems in the education industry

frequency	Neck [N=40.762]	Shoulders [N=40.750]	Arms/elbows [N=39.930]	Wrists/hands [N=39.982]	Back [N=41.024]
never	36.0%	39.1%	73.5%	68.6%	31.5%
Once, but for a short time	33.4%	29.6%	14.1%	17.9%	34.0%
Once, for a long time	4,3%	5,1%	3,4%	2,2%	6,9%
Multiple times, but for a short time	14,2%	13,1%	4,6%	6,1%	15,9%
Multiple times, for a long time	12,0%	13,0%	4,4%	5,2%	11,8%

Even though RSI in the upper extremities could still be a real problem to some people, the numbers of these complaints compared with those of neck, shoulders and back problems are a lot less common.

With what sort complaints have you called in sick at work, possibly more than one year ago?

• complaints neck, shoulders, arms, wrists, hands 2017 [N=41.649] 3,0%
• complaints neck, shoulders, arms, wrists, hands 2005 [N=22.693] 6,0%

According to the NEA over the last years, work related problems such as RSI have decreased in the Netherlands. This is probaly caused by the overall knowledge on working conditions since the mid- nineties, when RSI was a big thing. Nowadays problems such as RSI are prevented by guidelines on how to work with a correct posture and other preventives such as taking small breaks or using right equipment. Problems with the upper extremities can be solved quite easily most of the time by adjusting heights of the desk and office chair to the person. Neck, shoulder en back problems however are not so easily solved. These complaints can come from a lot of different sources, one of which is a faulty monitor.

RSI prevention: how to properly setup a computer monitor

In order to have the monitor in the optimal spot, the monitor needs to be right in front of the person. The upper side of the screen needs to be on the same height as the eyes. If the person cannot type blindly, or wears glasses, the screen should be placed a little bit lower (10 to 15 cm).

The optimal distance of the monitor would be an arm length. However, this optimal distance further depends on the width of the monitor and the size of the letters and symbols. A high resolution gives a better image, and thus smaller letters. The guideline for the optimal distance is 50 to 90 cm (dependent on the width of the screen).

The optimal angle between the eyes, the top of the screen and the bottom of the screen is 40 to 45 degrees (thus the angle between the top and the middle of the screen would be 20 to 22.5 cm). Whenever the eyes get lazy, they tend to focus in the distance. In this occasion the monitor should be placed a bit lower and a bit more to the back, in order to stay in the optimal position for the eyes.

Another important aspect is the amount of light on the monitor. In order to deal with incoming light from outside, the monitor can best be placed perpendicular to the window, at a distance of at least 3 meters. A brightness between 200 and 400 lux is considered to be optimal for the screen, but depending on the amount of light in the working place, the screen should be more or less bright. Also the contrast on the screen has to be as large as possible. Reflection of light (due to the sun, lamps, etc.) has to be reduced as much as possible.

https://www.jobat.be/extra/img/ideale-werkplek.jpg

http://www.rsi-alert.nl/images/werkhouding.gif

RSI prevention: how to use breaks to combat RSI

Physical ergonomics - the position of the monitor, height of the chair and desk etc. - are are often seen as the most important factor in preventing RSI, however it is certainly not the only factor that needs to be considered. [1] According Hanson et al. [2] the occurrence of RSI and the amount of hours working on a keyboard without a break are heavily associated with each other. So regular breaks from work to recover from physical and mental strain are vital to the prevention of RSI as well.

An argument that is often used against taking breaks is the loss of productivity due to that time not spend working. This is however not true, according to Galinsky et al. [3] discomfort in the right forearm, wrist and hand under a normal work schedule without breaks were eliminated without loss of productivity when regular scheduled breaks were introduced. According to Dababneh et al. [4] the productivity of workers was improved over the course of a workday due to taking breaks. In conclusion breaks don’t adversely affect productivity, it can even promote productivity.

However RSI is not only a problem in working environments, people at home are at risk of RSI as well with sufficient screen time in front of a computer. Break reminders can be used to remind users to take a break from their screen. These break reminders are often very customisable by the user and some can even learn when to suggest breaks based on work intensity and the user’s natural rest pattern. This causes reminders to come at the best possible moment instead of being very intrusive and annoying to the user. Example of break reminders are Workrave and BreakTimer.

Sensor Technology

Biometric authentication is the security process using the unique biological characteristics of a person to verify their identity. This form of authentication is used in information security, law enforcement, the daily life of individuals etc.

The most commonly used forms of biometric authentication are fingerprint scanning and facial recognition. [1][2] An example of fingerprint scanning can be found in smartphones, millions of people worldwide unlock their smartphone every day using fingerprint scanning. Ease of access and speed are the biggest reasons behind the increasing use of this form of biometric authentication. [3] An example of facial recognition can be found in Apple’s new flagship smartphone, the iPhone X. The iPhone X uses Face ID, a facial recognition system using more than 30000 infrared dots to accurately map and read faces for authentication and can adapt to the changes of the user’s face over time. [4]

There exist two types of fingerprint scanning techniques, offline scanning and live-scanning. Offline scanning uses fingerprints that are obtained on paper to create a digital image. [5] Live-scanning uses an electronic fingerprint scanner to obtain these fingerprints, this is also the scanning technique used for biometric authentication on smartphones. Live-scanning can be done using three kinds of sensors: optical, solid-state and ultrasound. Optical sensors are by far the most common and capture the fingerprints by measuring the differences in the amount of reflected light from each position in the fingertip and constructs an image based on the reflected light. Solid-state sensors measure differences in physical properties, for differences in electrical current in the different parts of the finger, the closer the finger is to the sensor the higher the capacitance is, in the finger to construct an image of the fingerprint. [6] Ultrasound sensors, as the name implies utilises ultrasound to map the fingerprint. Using pulses of ultrasound and measuring where the pulse of ultrasound is absorbed or bounced back an depiction of the fingerprint is created. [7]

A form of authentication that provides ease of access, speed and familiarity would be of great use to an RSI preventative AI. Biometric authentication has all these qualities, making it an ideal choice of authentication for the AI.

User Interface

(source text not yet available in our project folder)

Repetitive Strain Injury

RSI is thought to result from a continual risk of exceeding limits of “cumulative trauma load tolerance”. Painful stimulations also produce neuroplastic changes in the spinal and supraspinal nociceptive systems. RSI pain may be felt as a task-related response. (source 1)

In order to design a successful RSI prevention program, strategies have to be implemented to maintain compliance between the client and the program. Such a program can have an impact on the frequency of stretch breaks. (source 2)

RSI is a painful result of the inappropriate implementation of information technology in offices. However, RSI is also a management and organisational problem which should be approached in terms of an explicit health and safety policy. (source 3)

RSI is one of the main causes of lost working days. RSI requires a large amount of changes and preventive actions. (source 4)

The current strategy for RSI prevention breakdowns in the process is a risk factor for RSI. Technological systems are a potentially promising means for accident prevention, monitoring, detection and post-incident learning. (source 5)

Considering warnings, a computer warning led leads to the most correct position adjustments. The effect of the computer warning is caused by heightened attention for intervention. Warnings may be able to successfully replace educational brochures to produce behavioural changes. (source 6)

Dutch Arbo legislation

We are interested in researching RSI with respect to working life in the Netherlands. As such it seemed useful to include Dutch legislation regarding working conditions in the state-of-the-art literature study, as our design/prototype would have to comply with this legislation. In order to keep employees safe, the Dutch government introduced the working-conditions-legislation; the Arbowet (Arbowetgeving, n.d.). The Arbowet consists of three levels of legislation. The Arbowet itself offers the basis of the legislation, and contains general conditions for all working places. The Arbobesluit implements the core of the Arbowet into rules that both employers and employees should obey. The last level consists of the Arboregeling, which builds on the Arbobesluit and offers concrete regulation. The Arboregeling includes demands for work equipment.

Employers should minimize health risks for their employees, this can be done by replacing work equipment by more work-friendly alternatives (e.g. replace loud machinery with a quieter model). The working space should adapt as much as possible to the individual properties of the employees. The employers should also minimize the amount of monotonous work. Employees should also provide adequate training to their employees to for instance make better use of the work equipment.

Existing products

The Dutch RSI Vereniging is an association that focusses on providing information about the effects of RSI and how it can be prevented. They also provide a rundown of possibilities regarding the setup of the working environment in order to prevent RSI (inrichting werkplek, n.d.). Backshop is a distributor of ergonomic office space working equipment. By investigating their catalogue it appears that ergonomic office design focusses on using the following equipment:

• using ergonomically shaped mice and keyboards,
• using fully adjustable office chairs,
• using height adjustable desks, and
• mounting monitors to movable monitor arms

The Altwork Station (Specs, n.d.) integrates a monitor, desk and chair into a single unit. Every component is hinged, meaning that the user can adjust all components of the desk when the integrated seat is reclined. The user can specify 5 desks positions that can be stored, which allows the desk to change layout (e.g. from standing desk to sitting desk) with the use of a single button. Obutto is another company that distributes all-in-one ergonomic desks. Their design focusses on having a reclined chair, which spreads the user’s weight over their lower back and thighs. Their design features multiple movable platforms, which can be used to place a keyboard or mouse. These platforms can also be swiveled such that they can be used to make notes or draw. An early ergonomic desk design can be found in a patent by Nagy & Foris (1992), which describes a height adjustable desk with a rotating keyboard assembly. According to Nagy et al., their design allows the user to use their keyboard in the most comfortable way. They note that the cervical spine should be relaxed, and the computer monitor should be at eye level (Nagy & Foris, 1992, col. 7).

Motor Control

(source text not yet available in our project folder)

Relevant developments at Eindhoven University of Technology

(source text not yet available in our project folder)

Reference Lists

How long does it take for RSI to take effect?

1. Blatter, B.M. & Bongers, P.M. Duration of Computer Use and Mouse Use in Relation to Musculoskeletal Disorders of Neck or Upper Limb. 2002. Retrieved from: [1]
2. Hess, D. Employee Perceived Stress: Relationship to the Development of Repetitive Strain Injury Symptoms. 1997. Retrieved from: [2]
3. Clarke, A. RSI – Repetitive Strain Injury. 2018. Retrieved from: [3]

How big of a problem is RSI?

1. European Agency for Safety and Health at Work (2007). Work-related neck and upper limb disorders [fact sheet]. Retrieved from https://osha.europa.eu/en/tools-and-publications/publications/factsheets/72/view
2. Federatie Nederlandse Vakbeweging. RSI-klachten. (n.d.). Retrieved September 12, 2018, from https://www.fnv.nl/themas/veilig-en-gezond-werken/lichamelijke-belasting/rsi-klachten/
3. Koppes, L.L.J., de Vroome, E.M.M., Mol, M.E.M., Janssen, B.J.M., & van den Bossche, S.N.J. (2008). Nationale enquête arbeidsomstandigheden 2008: methodologie en globale resultaten. Retrieved from https://repository.tudelft.nl/view/tno/uuid%3A41ad2bd6-bba0-43b9-8a3e-e5678a48f6e5
4. Van Zwieten, M.H.J., de Vroome, E.M.M., Mol, M.E.M., Mars, G.M.J., Koppes, L.L.J., & van den Bossche, S.N.J. (2013). Nationale enquête arbeidsomstandigheden 2013: methodologie en globale resultaten. Retrieved from https://repository.tudelft.nl/view/tno/uuid:abe08063-3f54-452c-9919-a040084e63b9

RSI issues: upper extremity problems

1. Centraal Bureau voor de Statistiek, De Nederlandse Organisatie voor toegepast- natuurwetenschappelijk onderzoek. (2017) Nationale Enquête Arbeidsomstandigheden 2017 resultaten. retrieved from https://www.cbs.nl/-/media/_pdf/2018/16/nea2017%20tabel.pdf
2. Centraal Bureau voor de Statistiek, De Nederlandse Organisatie voor toegepast-natuurwetenschappelijk onderzoek. (2017) Nationale Enquête Arbeidsomstandigheden 2017 factsheet. Retrieved from http://www.monitorarbeid.tno.nl/dynamics/modules/SFIL0100/view.php?fil_Id=234
3. Centraal Bureau voor de Statistiek, De Nederlandse Organisatie voor toegepast-natuurwetenschappelijk onderzoek. (2017) Nationale Enquête Arbeidsomstandigheden 2017 methodisch rapport. Retrieved from http://www.monitorarbeid.tno.nl/dynamics/modules/SFIL0100/view.php?fil_Id=229
4. Centraal Bureau voor de Statistiek, De Nederlandse Organisatie voor toegepast-natuurwetenschappelijk onderzoek. (2005) Nationale Enquête Arbeidsomstandigheden 2005 methodisch rapport. Retrieved from https://www.cbs.nl/-/media/imported/documents/2013/27/2006-nea-rapport-2005.pdf

RSI prevention: how to properly setup a computer monitor

disclaimer: the writer of this section did not yet format his reference list to comply with the APA standard, this section will be updated when he has done so.

1. https://www.ergomotion.nl/beeldschermwerkplek-instellen/
2. http://www.ergonomiesite.be/computerwerkplek/
3. http://www.rsi-alert.nl/werkhouding.htm.php
4. https://www.jobat.be/nl/artikels/hoe-zit-je-gezond-achter-je-pc/
5. https://www.wijffels.nl/ergonomisch-werken/afstelling-werkplek
6. https://www.arbo-online.nl/rie/nieuws/2015/07/zicht-op-licht-bij-beeldschermwerk-10113442
7. http://www.sieso-ergo.eu/ergonomie/werkplek-inrichten-en-instellen

RSI prevention: how to use breaks to combat RSI

1. Taylor, K. (2002). The research on RSI and Breaks. [White paper]. Retrieved from https://www.wellnomics.nl/wp-content/uploads/2017/04/Wellnomics-white-paper-The-research-on-RSI-and-Breaks.pdf
2. Hanson MA, Donnan PT, Graveling RA, McLaren WM, Butler MP, Hurley JF, Kidd MW, Lancaster RJ, Prescott CA, Symes AM, Tesh KM (1999). Epidemiological and ergonomic study of occupational factors associated with syndromes of upper limb disorders in keyboard operators. [ONLINE] Available at: http://www.iom-world.org/pubs/IOM_TM9904.pdf [Last Accessed 12 September 2018]
3. Galinsky, T. L., Swanson, N. G., Sauter, S. L., Hurrel, J. J., & Schleifer, L.M. (2000). A field study of supplementary rest breaks for data-entry operators, Journal of Ergonomics, 43(5), 622-638.
4. Dababneh, A. J., Swanson, N., & Shell, R. L. (2001). Impact of added rest breaks on the productivity and well being of workers. Ergonomics, 44(2), 164-174. doi:10.1080/00140130121538

Sensor Technology

1. Spiceworks, Inc. (2018, March 12). Data Snapshot: Biometrics in the workplace commonplace, but are they secure? Retrieved September 9, 2018, from [4]
2. TheDigitalDentist. (2018, March 22). Biometric Authentication Seeing Wide Adoption In Businesses. Retrieved September 9, 2018, from [5]
3. Consumers Want Biometric Smartphones, and Love Fingerprint Scanning: FPC Study. (2017, October 02). Retrieved September 9, 2018, from [6]
4. Curtis, S. (2017, November 03). Face ID is the iPhone X's signature feature - here's how itworks. Retrieved September 9, 2018, from [7]
5. Nath, D., & Ray, S., & Ghosh, S.K. (2011, January). Fingerprint Recognition System: Design & Analysis. Paper presented at the International Conference on Scientific Paradigm Shift In Information Technology & Management.
6. Harris, T. (2002, September 24). How Fingerprint Scanners Work. Retrieved September 9, 2018, from [8]
7. Triggs, R. (2016, January 06). Ultrasonic fingerprint scanners: How do they work? Retrieved September 9, 2018, from [9]

User Interface

(source text not yet available in our project folder)

Repetitive Strain Injury

1. Sorgatz, H. (2002). Repetitive Strain Injuries. Forearm Pain Caused by Tissue Responses to Repetitive Strain.
2. Monsey, M., Ioffe, I., Beatini, A., Lukey, B., Santiago, A., James, A.B. (2003). Increasing Compliance with Stretch Breaks in Computer Users Through Reminder Software.
3. Khilji, N., Smithson, S. (1994). Repetitive Strain Injury in the UK: Soft Tissues and Hard Issues.
4. Maciel, R.H. (2000). RSI Prevention: A Brazilian Negotiated Program.
5. Teng, Y.C., Ward, J., Horberry, T., Clarkson, P.J., Patil, V. (2015). Retained Surgical Instruments: Using Technology for Prevention and Detection.
6. Visschers, V.H.M., Ruiter, R.A.C., Kools, M., Meertens, R.M. (2004). The Effects of Warnings and an Educational Brochure on Computer Working Posture: A Test of the C-HIP Model in the Context of RSI-relevant Behaviour.

Dutch Arbo legislation

1. Arbowetgeving. (n.d.). Retrieved from [10]
2. Arbeidsomstandighedenwet. (2018, January 1). Retrieved from [11]
3. Arbeidsomstandighedenbesluit. (2018, July 18). Retrieved from [12]
4. Arbeidsomstandighedenregeling. (2018, August 21). Retrieved from [13]

Existing products

1. Inrichting werkplek. (n.d.). Retrieved from [14]
2. Backshop producten. (n.d.). Retrieved from [15]
3. Specs. (n.d.). Retrieved from [16]
4. Obutto. (n.d.). Retrieved from [17]
5. Nagy, M.K., Foris, V.G. (1992). U.S. Patent No. US5174223A. Washington, DC: U.S. Patent andTrademark Office. Retrieved from [18]

Motor Control

Disclaimer These references will be replaced by precise references of relevant lectures within these courses as the design of the project evolves

1. TU/e 4DB00 (2016-2) Dynamics and control of mechanical systems course materials
2. TU/e 4GB20 OGO Robot-arm course materials

Relevant developments at Eindhoven University of Technology

Contactperson at TU/e: Eric van de Sande, Arbo- en milieucoördinator. Faculteit. WTB,EE

1. Konings, H. (2018). Contouren werkplekken worden zichtbaar in Atlas. Retrieved from [19]
2. Atlas facts and Figures. (n.d.). Retrieved from [Atlas-feiten2.pdf]