Tightening Torque Traceability: Bind It to the Bolt
By EVST Editorial Team · Reviewed by EVST Editorial Team · Method: written from a reviewed source-footage evidence map and published standards cited by designation; no performance, tolerance, cycle-time or deployment figure is asserted. · · Editorial policy · Corrections policy · Terms
Direct answer: Verify that the part is located by the fixture rather than found by the robot, that the tightening sequence comes from the process document, and that every torque and angle record binds to a specific part and a specific position on it. A batch average is a statistic, not a traceable record. The data path that makes the binding possible is not visible in any video and has to be checked directly.
Who this is for: Written for engineers and quality staff specifying automated tightening where results have to be recorded and traced.
Scope: This EVST guide covers part location, tightening sequence and the data path that makes a torque record traceable. It does not quote torque values, capability indices or cycle times.

Location comes before torque
In the reference footage the block arrives on the line and stops on a fixture. Its position is set by that fixture, not found by the robot.
If location is not repeatable, accurate torque data still cannot be tied to the right position on the part. Repeatability of the fixture and the presence detection belong in the specification.
Sequence is a process requirement
The robot works each position in order. Tightening order affects how the joint faces seat, so it usually comes from the process document rather than from the programmer.
Target torque and the angle window for each position belong in the same document. The station executes it; it does not decide it.
According to ISO 5393:2017, the capability of a power tool for threaded fasteners is established against a stated joint, which is why a torque figure without its joint definition is not evidence.

What tightening torque traceability actually requires
Recording torque and angle is the first step and the easy one. Most controllers do it.
Traceability requires the record to bind to a specific part and a specific position on that part. A batch average is a process statistic, not a traceable record.
That binding is what makes it possible to answer a question about one assembly months later, which is usually why the requirement exists in the first place.
The data path is not visible in any video
No footage can show where the data leaves the controller, in what format it arrives, whether it carries a part identifier, or how long it is retained.
Those four questions are the ones to put to a supplier directly, and the answers belong in the proposal rather than in a later integration phase.
| Element | Why it matters | Where it is verified |
|---|---|---|
| Part location | Ties a result to a position on the part | Fixture repeatability and presence detection |
| Tightening sequence | Affects how joint faces seat | Process document, not the robot program |
| Torque and angle record | The measurement itself | Controller output format |
| Part identifier binding | Turns a number into a traceable record | Interface between controller and quality system |
| Rework capture | Closes the chain for out-of-window results | Station sequence and record retention |
Rework and out-of-window results
A traceable station has to record the results that fall outside the window as well as those inside it, and it has to make clear what happened to that assembly afterwards.
How rework is captured is often the weakest link in an otherwise complete data chain.
Collaborative operation does not remove the assessment
The arm in the footage is a collaborative type working close to an assembly line. That does not by itself make the station safe without an application risk assessment.
ISO 10218-2:2025 covers integration and ISO 12100:2010 the assessment; the tool, the part and the layout all enter into it.
According to ISO 10218-2:2025, collaborative operation is a property of the application rather than of the arm, so the tool, the part and the layout all enter the assessment.
According to ISO 12100:2010, that assessment covers the whole station as installed, including the fixture and the conveyor it sits on.
In practice an application review collects the project inputs first: the joint list, the record format and the fixture. Selection of tool and controller follows from that evidence, and acceptance is agreed on what the records have to prove.
Frequently asked questions
Is recording torque the same as traceability?
No. Recording produces a number. Traceability requires that number to bind to a specific part and a specific position, and to remain retrievable later.
Does the robot decide the tightening sequence?
It should not. Sequence affects how joint faces seat and normally comes from the process document.
What about results outside the window?
They have to be recorded too, along with what happened to that assembly. Capturing rework is often the weakest part of a data chain.
What should we send to get a useful answer?
The joint list with target torque and angle windows, and the record format your quality system needs.
Project inputs for an application review
To have the data path checked against your requirement, send:
- joint list with target torque and angle windows
- record format and retention your quality system requires
- part location method and fixture repeatability
- cycle requirement and variant mix
Send the joint list and the record format you need and the data path can be checked against it. Related reading: industrial robot product range, assembly line integration.