RULA (Rapid Upper Limb Assessment) is a screening tool for postural risk in the upper body, developed by McAtamney and Corlett at the University of Nottingham in 1993. It is designed for work where the load on the neck, trunk and upper limbs comes from posture rather than from lifting heavy weights — assembly, machine operation, laboratory and desk-based tasks. Motionprint Ergo implements it in full — automatically extracting every joint angle from your motion capture data and scoring every frame of the recording. This article explains how the method works, what each scoring component represents, and how to interpret your results.
1 Overview
RULA produces a single RULA score from 1 to 7 for each assessed posture task. The score determines the risk level across four categories:
| RULA score | Risk level | Interpretation |
|---|---|---|
1 – 2 | Negligible risk | Acceptable — the posture is acceptable if it is not maintained or repeated for long periods. |
3 – 4 | Low risk | Further investigation needed, and changes may be required. |
5 – 6 | Medium risk | Changes are needed soon. |
7 | Very high risk | Changes are needed now. |
RULA is deliberately a screening method. A high score identifies a task that warrants a closer look; it does not by itself quantify injury probability, and it is not a substitute for a full biomechanical analysis. Its strength is speed and coverage — it can flag problem tasks across a whole workplace quickly.
Every joint angle RULA needs — upper arm, lower arm, wrist, neck and trunk flexion, plus the twist, side-bend and abduction adjustments — is derived directly from the Xsens MVN body model. No goniometer, no visual estimation, and no frame-by-frame stopwatch work.
2 How the Score Is Built
RULA scores two body regions separately and then combines them through a lookup table. Group A covers the arm and wrist; Group B covers the neck, trunk and legs.
Each group is built from its own set of components, each scored from a joint angle and then adjusted for secondary conditions:
Upper arm, lower arm, wrist position and wrist twist. The four component scores are combined in Table A into a single posture score for the limb.
Neck, trunk and legs. The three component scores are combined in Table B into a second posture score.
One point when the posture is mainly static or the action is repeated. Evaluated separately for each group, and added to that group's total.
Zero to three points from the weight handled and how the force is applied. The same value is added to both group totals.
Two consequences follow from this structure, and both matter when you read a report. First, the Force/Load points are applied twice — once to each group — so a single change to the load input can move the final score by more than one point. Second, because Table C is a lookup rather than a sum, the relationship between the group totals and the final score is non-linear: improving an already-poor posture by one point sometimes changes nothing, and sometimes changes the risk level.
3 Group A — Arm and Wrist
Group A scores the working limb. Each component starts from a base score set by a joint angle, then gains or loses adjustment points for secondary conditions. All four are combined in Table A.
| Component | Scoring | Source |
|---|---|---|
| Upper arm (shoulder flexion/extension) | 1 to 4 — within 20° of neutral scores 1; 20–45° scores 2; 45–90° scores 3; beyond 90° scores 4. Extension beyond 20° is treated the same as the equivalent flexion. Adds +1 if the arm is abducted beyond 45°, +1 if the shoulder is raised beyond 10°, and subtracts −1 if the arm is supported or the person is leaning on it — floored at 1. | Derived automatically from motion capture; arm support entered per task |
| Lower arm (elbow flexion) | 1 or 2 — the comfortable 60–100° range scores 1; anything outside it, whether straighter or more closed, scores 2. Adds +1 if the hand works across the body's midline or out beyond its silhouette. | Derived automatically from motion capture |
| Wrist (flexion/extension) | 1 to 3 — within 5° of neutral scores 1; 5–15° scores 2; beyond 15° scores 3. Scored by magnitude, so bending up and bending down are treated alike. Adds +1 if the wrist is deviated beyond 10° toward the thumb or little finger. | Derived automatically from motion capture |
| Wrist twist (forearm rotation) | 1 or 2 — rotation near the middle of its range scores 1; rotation at or near either end scores 2. The range is not symmetric, so the boundaries differ: beyond 45° palm-down, or beyond 30° palm-up. | Derived automatically from motion capture |
The −1 for a supported arm applies to the upper arm component and the result is floored at 1. If the upper arm already scores 1, adding support cannot take it to 0 — RULA has no zero. This is why a report can show a supported arm and a raised shoulder and still display 1 for the upper arm.
4 Group B — Neck, Trunk and Legs
Group B scores the posture supporting the working limb. The same base-plus-adjustment pattern applies, and the three components are combined in Table B.
| Component | Scoring | Source |
|---|---|---|
| Neck (flexion/extension) | 1 to 4 — 0–10° of flexion scores 1; 10–20° scores 2; beyond 20° scores 3. Any meaningful extension — the head tipped back beyond 5° — scores 4 outright, the worst available score, because it loads the cervical spine far more than forward flexion. Adds +1 for twisting beyond 20° and +1 for side bending beyond 10°. | Derived automatically from motion capture |
| Trunk (flexion/extension) | 1 to 4 — within 5° of upright scores 1; 5–20° scores 2; 20–60° scores 3; beyond 60° scores 4. Scored by magnitude, so leaning back falls in the same band as the equivalent forward bend. Adds +1 for twisting beyond 10° and +1 for side bending beyond 10°. | Derived automatically from motion capture |
| Legs (support and balance) | 1 or 2 — scores 1 when the legs and feet are supported and the weight is evenly balanced, and 2 when they are not. Derived from the body's centre of mass relative to the base of support; any frame where both feet are not on the ground is forced to 2, since standing on one foot is not evenly balanced whatever the centre of mass is doing. | Derived automatically from motion capture |
Because extension jumps straight to 4, a task that involves looking upward — overhead assembly, inspecting a raised surface — can produce a markedly higher score than the same task performed at chest height, even when nothing else changes. This is intended behaviour in the method, not an artefact of the measurement.
5 Muscle Use and Force
Posture alone does not determine risk. A posture held for a long time, or repeated many times a minute, is more demanding than the same posture adopted briefly, and a load in the hands is more demanding than an empty hand. RULA captures both with two additions applied to each group total.
Muscle use
Muscle use adds +1 when either of two conditions is met — and both are determined automatically from the recording rather than entered by the assessor.
| Condition | How it is obtained |
|---|---|
| Static posture | The joint angles for that group are tracked across the posture task. If they stay within a narrow band for longer than a minute — allowing for natural sway rather than demanding perfect stillness — the posture counts as mainly held. |
| Repetitive action | The number of posture tasks is divided by the trimmed duration of the recording in minutes. At four or more occurrences per minute the action counts as repeated. This is a report-level figure, so it applies to every task in the assessment. |
There are two separate muscle use scores, not one: the condition is evaluated independently for the arm and wrist group and for the trunk, neck and legs group, and each result is added to its own group total. A worker can hold the trunk still while the arms repeat a cycle, and the two scores capture exactly that — one group can attract the point while the other does not.
Force / load
The force component reflects the weight handled and how the force is applied. The weight is entered per posture task, as is a flag for force applied with shock or a rapid build-up.
| Condition | Points |
|---|---|
| Less than 2 kg, intermittent | 0 |
| 2 – 10 kg, intermittent | +1 |
| 2 – 10 kg, static or repeated | +2 |
| More than 10 kg, or a static / repeated load | +3 |
| Any load applied with shock or a rapid build-up of force | +3 |
The shock condition overrides the weight bands entirely. A 3 kg load applied with a sudden impact scores 3, not 1 — and because the points apply to both groups, setting that single flag can move the final RULA score by two points.
Shock loading describes how force is applied — hammering, snapping a part into place, absorbing a power tool's torque reaction. It has no reliable signature in motion data: bracing against a torque reaction barely moves the hand, while a fast unloaded reach accelerates hard. Rather than invent a threshold, Motionprint Ergo asks the assessor, exactly as the published RULA worksheet does.
6 The RULA Score
The two group totals meet in Table C, which returns the final score from 1 to 7:
| Table B total — neck, trunk and legs | |||||||
|---|---|---|---|---|---|---|---|
| Table A total | 1 | 2 | 3 | 4 | 5 | 6 | 7+ |
| 1 | 1 | 2 | 3 | 3 | 4 | 5 | 5 |
| 2 | 2 | 2 | 3 | 4 | 4 | 5 | 5 |
| 3 | 3 | 3 | 3 | 4 | 4 | 5 | 6 |
| 4 | 3 | 3 | 3 | 4 | 5 | 6 | 6 |
| 5 | 4 | 4 | 4 | 5 | 6 | 7 | 7 |
| 6 | 4 | 4 | 5 | 6 | 6 | 7 | 7 |
| 7 | 5 | 5 | 6 | 6 | 7 | 7 | 7 |
| 8+ | 5 | 5 | 6 | 7 | 7 | 7 | 7 |
Table C, shaded with the same risk-band colours the app uses on the score overview.
Table A totals of 8 or more and Table B totals of 7 or more are clamped to the last row and column — beyond that point the method does not distinguish further, because the task already needs immediate attention.
Laid out this way, two properties of the method become visible. The acceptable corner is small: only four of the fifty-six combinations score 1 or 2, so most real tasks land in a band that calls for some kind of follow-up. And the bands step diagonally rather than smoothly, so a one-point improvement in a single group frequently leaves the final score untouched — of the forty-nine ways to lower a Table A total by one point, twenty-eight return the same score as before. Dropping a Table A total from 4 to 3 against a Table B total of 3, for example, leaves the score at 3. Moving a task into a lower band usually takes a change that reduces both groups, or a large enough change in one to cross a boundary.
It is worth understanding what the score does and does not tell you. A score of 7 means the posture demands immediate change; it does not mean the task is seven times worse than a score of 1. The scale is ordinal. Comparing scores between two tasks tells you which needs attention first, not by how much.
Equally, a low RULA score is not a guarantee of safety. RULA assesses posture, muscle use and force in the upper body. It does not assess vibration, thermal conditions, contact stress, or the cumulative effect of a full shift. For lifting tasks specifically, a load-focused method such as NIOSH or KIM-LHC is the appropriate tool — RULA and those methods answer different questions and are complementary rather than interchangeable.
7 Implementation in Motionprint Ergo
Motionprint Ergo implements the full RULA method in a dedicated processing pipeline that applies the published scoring bands to the motion capture data frame by frame, across all defined posture tasks.
Calculation workflow
Upper arm, lower arm, wrist, neck and trunk angles, along with the twist, side-bend, abduction and shoulder-raise measurements and the body's centre of mass, are extracted automatically from motion capture data. Manual inputs — load weight, arm support, shock loading, and the trim range defining the active portion of the recording — are entered in the assessment setup panel and passed to the processing pipeline.
Each component is scored for every frame of every posture task, with its adjustments applied. Because Group A describes one limb at a time, the arm and wrist components are scored separately for the left and right sides. Muscle use is determined per group from how long the posture is held and how often tasks occur.
The Group A and Group B posture scores are formed from Tables A and B, muscle use and force points are added to both, and Table C returns the RULA score for that frame. For each frame the higher-scoring side is taken, and each task is then reduced to its peak frame — the highest score reached — which is the result the report presents.
Results are presented across three report tabs. Score overview gives you everything a completed RULA worksheet would: every component score with its adjustments, the three scoring tables with the cells used, and the final result. In-depth analysis is for exploring how each measurement behaves over time, so you can pinpoint which postures and moments drive the score. Findings lets you document your own observations.
Next steps: to run your first assessment, see Running a RULA Assessment. To interpret a completed report in detail, see Reading the RULA Report Results, and to change the defaults applied to every assessment, see RULA Report Settings.