The buyer task and decision boundary

An application dossier should connect the original machine, the proposed component and the task that will demonstrate the intended result. The most useful plans are often narrower than a complete repair promise: establish a drive combination, review a tool and harness interface, or reconcile the evidence required by the receiver. Each plan needs explicit inputs, review branches and artefacts. This makes a sourcing enquiry concrete without inventing an available product or a service job already performed.

Technical context: End effectors: document the task and failure behaviour

A tool replacement affects more than the flange. List the complete attached mass, the centre-of-gravity information required by the robot maker, the workpiece contact surfaces and the utility connections. Explain what the tool does if power, air or communication disappears. The tool-close command, the tool status bit and evidence that a workpiece is retained are three different things. A useful trial includes representative objects and failed attempts, not only a carefully positioned sample. When an adapter, finger or protection layer changes, preserve the version so later reviewers know which physical arrangement produced the result.

A second condition to preserve: Cables and connectors: follow the complete moving route

A harness is a connection map and a moving assembly. Connector keying does not establish the pin assignment, and wire colour is not an approved pinout. Identify power, feedback, communications, shielding and reference connections from applicable drawings. Include the route through the complete intended travel, the strain relief and the replacement access. A cable can pass a static continuity check yet fail when routed through motion; conversely, a controller symptom does not prove the cable is defective. Functional checks must follow the equipment procedure and use the appropriate personnel, without experimenting with unidentified live pins.

Choose a task-sized application package

In a parts or tooling dossier, the robot, tool and receiver evidence must remain connected. A useful application plan is narrow enough that the parties can recognise its inputs and outputs. Start with a real decision, describe the operating or documentary context and identify the people who provide evidence. Separate preparation, review, actual implementation and acceptance. These stages need not be purchased together. Making the boundary explicit helps a buyer commission the missing work without interpreting a planning document as a completed installation or a measured result.

Sequence the work with observable exits

In a parts or tooling dossier, the robot, tool and receiver evidence must remain connected. For each stage, write the required input, the action, the artefact and the condition for continuing. A document received is not automatically reviewed, a command acknowledged is not necessarily a physical result, and a calculated value is not a measurement. If a condition cannot be confirmed, return to the relevant input rather than advancing because the schedule has moved forward. Preserve unsuccessful attempts and explain any change to the sequence.

Review three evidence packages separately

In a parts or tooling dossier, the robot, tool and receiver evidence must remain connected. The three application scopes below address different problems. They may share identities and documents, but each should have its own question, deliverables and unresolved items. Keep common data in a versioned input list and reference it consistently. Where one package depends on another, state the dependency and avoid counting the same evidence twice as if it were independent confirmation. A receiving team should understand which decision each package can support.

Plan acceptance while the input is still being agreed

In a parts or tooling dossier, the robot, tool and receiver evidence must remain connected. Acceptance should specify how another person can inspect the artefacts and understand the tested or reviewed conditions. Define the expected and actual result, the treatment of failures, the observation boundary and responsibility for open items. Do not invent a universal pass count. The package is useful when it leaves a reproducible route to the next decision, including an honest return-to-review branch when the available evidence does not support continuation.

Apply the worksheet to this domain

Use the input sheet to describe robot serial identity and installed part number, then connect drawing and pinout revision and verification task to the selected application package. For every stage, identify the supplied input, the planned artefact and the condition for moving on. Make the dependency on another package explicit. A receiver should be able to see where preparation ends and actual implementation begins. Preserve failed and not-covered branches in the plan; do not turn a sequence of planned actions into a claim that the task has already been performed.

Three task-sized application scopes

Drive replacement evidence pack addresses this problem: A drive-related alarm exists without a complete revision map. Its scope is controller, drive, motor and feedback combination. Review its inputs, planned artefacts and open conditions separately from the other packages.

Tool and moving-harness review addresses this problem: A mating connector or flange is being treated as full compatibility. Its scope is tool contact, utilities, pinout and complete travel route. Review its inputs, planned artefacts and open conditions separately from the other packages.

Post-replacement handover plan addresses this problem: The receiver lacks a clear task and observation boundary. Its scope is restoration scope, original trigger and receipt inspection. Review its inputs, planned artefacts and open conditions separately from the other packages.

Prepare the next review with original evidence

For the next planning review, choose one of the three packages and identify its input owner and receiving role. Bring the source documents and the task context, then agree stage exits, artefacts and treatment of missing conditions. Do not ask every package to prove the same result. Use the references below as technical background, and record which definitions apply to this task. The output should be a sequence that can be reviewed and commissioned in bounded steps.

[S1] Harmonic Drive Engineering Data; [S2] UR-Series Service Manual.

Primary-source reading

Harmonic Drive Engineering Data
https://www.harmonicdrive.net/_hd/content/documents1/FBB_DifferentialGear.pdf

UR-Series Service Manual
https://www.universal-robots.com/manuals/EN/PDF/Shared/service-manual_ur-series/Service_Manual_UR-Series_en.pdf