The OWAS Method: How to Analyze Working Postures With the Ovako System

What is the OWAS method?

The OWAS method, short for Ovako Working Posture Analysing System, is one of the oldest observational methods for assessing working postures. It was developed at Ovako Oy, a Finnish steel producer, and published by Karhu, Kansi and Kuorinka in Applied Ergonomics in 1977.

The original paper describes two parts: an observational technique for coding postures, which work-study engineers could use after a short training period, and a set of criteria for redesigning work methods and workplaces. The criteria were based on evaluations by experienced workers and ergonomics experts, with the main emphasis on the discomfort caused by working postures.

According to a 2017 review of OWAS applications, the method was first used to evaluate workload in the repair of smelting furnaces and began with 72 postures identified by photographing work at the company.

This post is general information, not medical advice. Anyone with persistent pain should see a clinician.

How does OWAS posture coding work?

OWAS describes every observed posture with a four-digit code: one digit each for the back, arms, legs and load. The codes below follow the standard OWAS classification as reproduced by Esen et al. (2015).

Back (first digit)

Code	Back posture
1	Straight
2	Bent forward or backward
3	Twisted or bent sideways
4	Bent and twisted, or bent forward and sideways

Arms (second digit)

Code	Arm posture
1	Both arms below shoulder level
2	One arm at or above shoulder level
3	Both arms at or above shoulder level

Legs (third digit)

Code	Leg posture
1	Sitting
2	Standing with both legs straight
3	Standing with the weight on one straight leg
4	Standing or squatting with both knees bent
5	Standing or squatting with one knee bent
6	Kneeling on one or both knees
7	Walking or moving

Load or force (fourth digit)

Code	Load
1	10 kg or less
2	More than 10 kg but less than 20 kg
3	More than 20 kg

Four back postures, three arm postures, seven leg postures and three load classes give 4 x 3 x 7 x 3 = 252 possible combinations. A worker bending and twisting, with both arms low, standing on both straight legs, holding 12 kg, would be coded 4-1-2-2.

What are the OWAS action categories?

Each four-digit code is assigned to one of four action categories using the OWAS classification table. The categories, as summarized in the 2017 OWAS review and Esen et al.:

Category	Meaning	Action
1	Normal posture with no harmful effect	No corrective action needed
2	Posture with some harmful effect	Corrective action in the near future
3	Posture with harmful effect	Corrective action as soon as possible
4	Posture with very harmful effect	Corrective action immediately

The full lookup table that maps all 252 codes to categories is part of the original method. Use a published OWAS chart or software to assign categories rather than judging them by eye.

How do you run an OWAS study?

The OWAS method was designed around work sampling: observing at regular intervals to estimate how much of the working time is spent in each posture. The process the review describes has four steps:

1. Observe the work tasks. Decide the observation period and sampling interval, and make sure the sample covers the full range of tasks.
2. Code the postures. At each sampling point, record the four-digit code.
3. Assign risk categories. Look up each code's action category.
4. Propose corrective actions. Focus on the postures and tasks that land in categories 3 and 4, and on those that take up the most time.

Video makes this far easier. You can sample frames at a fixed interval after the fact, check difficult postures, and have a second person code the same frames to test consistency.

The output is usually a distribution, for example "18% of observed postures were category 3 or 4, mostly during floor-level picking." That kind of statement is what makes OWAS useful for prioritizing across a whole job or shift.

What are the strengths of the OWAS method?

• Simple and quick to learn. The review notes it can be used by people from health, engineering and industry without specialized training.
• Consistent between raters. In Kee's 2022 comparison, OWAS had the highest inter- and intra-rater reliability of OWAS, RULA and REBA, which Kee links to its simplicity.
• Built for whole-shift sampling. Because each observation is just a code, you can log hundreds of samples and see how exposure is spread across a day.
• Detailed leg coding. Seven leg categories capture kneeling, squatting and walking, which RULA cannot separate.

What are the limitations of OWAS?

The same reviews are clear about the gaps:

• No neck, elbow or wrist coding. OWAS cannot pick up a bent neck or a deviated wrist.
• Coarse shoulder coding. Arms are only "below" or "at or above" shoulder level.
• No left/right distinction. The arm code does not record which side is raised.
• No repetition or duration factor per posture. Time in posture only shows up through the sampling distribution.
• Tends to rate risk lower. In Kee's review, RULA rated postures as more stressful than OWAS in 9 of 10 comparison studies, and REBA rated them as more stressful in 10 of 13.

These limits are why the OWAS method is often paired with a more detailed tool for the worst postures it finds.

How does the OWAS method compare with REBA and RULA?

Feature	OWAS	REBA	RULA
Body parts	Back, arms, legs	Trunk, neck, legs, upper/lower arms, wrists	Upper/lower arms, wrists, neck, trunk, legs
Output	4-digit code, 4 action categories	Score 1-15, 5 action levels	Score 1-7, 4 action levels
Typical use	Work sampling across a shift	Single worst posture, whole body	Single worst posture, upper body
Load	3 classes	0-2 plus shock	0-3

A practical workflow is to use OWAS to find which tasks spend the most time in categories 3 and 4, then score the worst postures in those tasks with REBA. Our RULA vs REBA comparison covers which of the two detailed tools to pick.

Can OWAS be automated?

Researchers have tried. The 2017 review lists studies applying OWAS with video, image processing, virtual reality and digital human models. The idea is the same one behind modern pose-estimation tools: estimate body segment positions automatically, then apply the coding rules.

ViveLab does not currently produce OWAS codes. Its posture analysis from video estimates REBA and RULA scores from a phone video or photo, with processing on your device. For an OWAS-style workflow, you can use OWAS to sample the shift, then run the worst frames through the free REBA calculator or the AI analysis, with a trained assessor reviewing the results.

Next steps

1. Film a representative stretch of the shift and sample frames at a fixed interval.
2. Code each frame with the four-digit OWAS code and assign action categories from a standard OWAS chart.
3. Take the category 3 and 4 postures and score them in detail with the REBA calculator.

For an overview of where OWAS sits among other methods, see our guide to ergonomic risk assessment tools.