From OSHWiki
Jump to: navigation, search

Richard Graveling, Institute of Occupational Medicine, Edinburgh, UK


There are many sources of information that go into considerable detail regarding anthropometry.  For example this article in Wikipedia[1] presents a general overview of the topic, summarising different approaches to measurement of the human form and different fields of application, while other sources present a more focussed and detailed description of its use in workplace design[2].  This present article presents a brief introduction to its use in ergonomic design.

As it says in the main article on ergonomics:
“Anthropometry is the science of measurement of the human body. It can be applied to OSH to ensure that workers have sufficient space to perform their tasks, that they can reach necessary equipment, tools and controls, that barriers keep them out of reach of hazards, and that working postures can be optimised for the range of people using them.”
The need for anthropometry and anthropometric data in ergonomic design stems from the fundamental fact of human variability – people of all parts of the world differ from one another in many ways (not just physically) – and takes, as a particular focus, the variability in size and shape of the human body.  ISO 7250[3] provides further detail regarding the most important physical anthropometric dimensions for workplace design.

Although many different factors influence that variability; race and gender are two of particular practical importance.  For example when a machine tool is designed for one national population it may not fit well when sold and installed for use in another country with a different population.  For example, elbow heights for a Belgian male adult population (5th – 95th percentile) range from 1047 mm to 1217 mm.  In contrast, the equivalent range for Italian males is 1004 mm to 1172 mm.  Elbow height is an important dimension for establishing the height for working surfaces and a difference of approximately 40 mm can make a substantial difference.  With some other countries the differences can be even more pronounced (e.g. 5th percentile elbow height for Japanese male adults is 975 mm)[4].

Another example, although a diminishing problem – is the design of Personal Protective equipment (PPE), often designed by men with men in mind, that has proved too large for many women.  Where good fit is essential for effective protection (e.g. much respiratory protective equipment) this difference can have serious consequences. Even where a good fit is less important for effective protection a poor fit can have an adverse effect on comfort, which may reduce the willingness of workers to wear them.  Simply scaling the design down does not always solve the problem – because women often have different facial shapes, not just sizes.

The increased mobility of people between countries and ongoing trends to increase the employment of women means that such factors are of growing relevance.

The normal distribution

Many human physical dimensions follow a ‘Normal distribution’, that is to say they reflect a symmetrical pattern (often described as a bell-shaped curve) as shown in Figure 1[5].

Figure 1: Curve of the Normal distribution (from wikipedia)

The peak of the curve represents the average or mean value (and also, if genuinely symmetrical, the median and mode – the central and commonest - values) with the values to either side of this peak featuring less frequently amongst the population the dimension is drawn from.  It will be seen from this that, the further you get away from the centre, the less frequently such values appear in the population under scrutiny.

From an ergonomics perspective, ISO 26800, the core standard on ergonomics principles, states:
“In most instances, the use of average values is not an adequate way of accommodating the range of values to be found associated with a particular characteristic.”[6]
Figure 1 includes markers reflecting a number of ‘standard deviations’[7] from the mean value as a measure of the variability around the mean.  For dimensions where the standard deviation is small, the equivalent curve will be narrower but taller (and wider and lower for a large standard deviation).

It will be seen from Figure 1 that slightly over two-thirds of the values (68%) are no more than one standard deviation from the mean.  Extending the range to two standard deviations accommodates 95% of the values.  Note how increasing the range still further, to three standard deviations, adds less than 5% to the total (4.7%).  In some instances, rather than work in standard deviations, reference is made to ‘percentiles’ – reflecting the percentage of the population encompassed.

Static anthropometry

As a start point for many designs, the variability (within the design population) of the size (length, width, etc.) of relevant body parts should be taken into account.

ISO 26800 explains:
“In ergonomics, the variation within the target population is commonly accounted for by using the 5th and/or 95th percentiles of important design characteristics (e.g. body size, visual abilities, literacy), with the intention of accommodating at least 90% of the target population. In some circumstances, a different percentile range is used. For example, in many safety-related applications, the 1st and 99th percentiles are used.”
To express that another way, designs drawn up to this ‘rule’ would accommodate all but the smallest 5% (the 5th percentile) and the largest 5% (the 95th percentile).  As a good rule of thumb in designing workplaces, elements that are needed to be reliably reached are placed so that smaller members of the population can reach them (usually adopting the smallest 5% as a target).  Those elements where the aim is to ensure sufficient clearance, are designed with the larger members of the workforce in mind.  As ISO 26800 explains, where considered necessary a wider range of percentiles than 5% - 95% might be used.

Failure to adopt such an approach can lead to safety and health problems.  For example, placing a safety guard at a distance away from a cutting tool equivalent to average arm length would mean that half the workforce could still reach the tool over the guard – perhaps to clean it – with obvious potential injury concerns.

As an example of the mis-use of a static anthropometric dimension, an industrial plant was seen where control valves on a pipe were placed at a height of 2110 mm. Enquiries indicated that the design engineers had used data that included average vertical reach.  Fifth percentile vertical reach is 1960 mm – meaning that, for some of the workforce, the valves were 150 mm above their hands when they reached up to close or open them.

There are many sources or databases of relevant dimensions, some are available as paper documents, others are electronically available.  Some are integrated into design packages that allow the designer to automatically incorporate differing anthropometric percentiles (or different design populations) into workplace designs.

Dynamic anthropometry

Physical (body) dimensions are not the only anthropometric characteristics taken into account in ergonomics design.  These are examples of ‘static anthropometry’.  ‘Dynamic anthropometry’ (e.g. ranges of movement) can also be important.  Thus, work activities that make use of too wide a range of a joint’s range of movement can lead to an increased risk of musculoskeletal disorders in that joint.  One rough guide in this regard is to remain within the middle 50% of the range of movement, noting that the range might not be symmetrical around the neutral or resting position.  However, it is important to recognise that there can be exceptions to this rule which shouldn’t therefore be followed without question.

An example of the application of anthropometry to workplace design

As an example of the application of static anthropometry to design, consider the simple example of the height of a work surface for manually handling products onto. Ideally this should be between knuckle and elbow height.  Any less than knuckle height then some workers will need to bend down whenever lifting to or from that bench, possibly leading to back problems.  Heights greater than elbow height may require the arms to be raised, or the back to be bent back (arched), again leading to a risk of injury.

According to one electronic commercial database[4], knuckle height ranges from 665 mm for a 5th percentile British working age woman to around 890 mm for a 95th percentile British working age man.  From the same database, elbow height ranges from 920 mm to around 1220 mm for the same range of workers (values rounded to 5mm).  Allowances should be made for footwear as measurements are usually obtained from barefoot subjects, adding 25-45 mm to these figures.  Adding 25 mm for footwear it will be seen that benches less than 915 mm high will require taller men to stoop, while benches higher than 935 mm will entail some lifting above elbow height for shorter women.  Adopting a height between these two will therefore provide a suitable height for such tasks. Note that this gives a slightly larger coverage of the workforce than would be accommodated by taking the 5th and 95th percentile values for a mixed (male and female) population.

ISO 14738[8]:
“establishes principles for deriving dimensions from anthropometric measurements and applying them to the design of workstations at non-mobile machinery.”
“It specifies the body’s space requirements for equipment during normal operation in sitting and standing positions.”
This standard requires the user to determine the necessary dimensions based upon the selection and application of anthropometric data appropriate for the intended user population.  Others make assumptions about the user

population and present actual dimensions. 

For example, ISO 13857 specifies safety distances to prevent hazard zones being reached by upper and lower limbs, based on assumptions about those who it might apply to.  As the scope states:
“This International Standard covers people of 14 years and older (the 5th percentile stature of 14 year olds is approximately 1,400 mm). In addition, for upper limbs only, it provides information for children older than 3 years (5th percentile stature of 3 year olds is approximately 900 mm) where reaching through openings needs to be addressed.”[9]
However, it adds as a rider:
“Because safety distances depend on size, there will be some people of extreme dimensions who will be able to reach hazard zones even when the requirements of this International Standard are complied with.”
Material from standards such as these enables anthropometric dimensions, and variation in those dimensions to be taken into account in designing workplaces, helping to ensure safe and healthy workplaces for those who use them.


  1. Wkipedia article on 'anthropometry'. [1]
  2. Pheasant, S. & Haslegrave, C.M., Bodyspace: Anthropometry, Ergonomics and the Design of Work, 3rd edition, Taylor & Francis, London, 2006.
  3. ISO 7250: Basic human body measurements for technological design. Geneva: ISO
  4. 4.0 4.1 PeopleSize, Open Ergonomics Ltd
  5. Wikipedia article 'Normal distribution'
  6. ISO 26800: Ergonomics -- General approach, principles and concepts. Geneva: ISO
  7. Wikipedia article on 'Standard deviation'
  8. ISO 14738: Safety of machinery. Anthropometric requirements for the design of workstations at machinery. Geneva: ISO
  9. ISO 13857: Safety of machinery — Safety distances to prevent hazard zones being reached by upper and lower limbs. Geneva: ISO