Every systems engineer has watched it happen, and most of us have done it. A first Capella model begins as an honest Operational Analysis: a few actors, a capability, an activity or two. Then someone asks how the system will support that activity, and a system function appears. The function wants a home, so a logical component turns up. The component obviously runs on something, so a physical box gets drawn “just as a placeholder.” By the end of the afternoon the model touches every Arcadia layer, finishes none of them, and the review scheduled for Friday has nothing it can accept — because nothing on the desk said where this increment was supposed to stop.
The model is not wrong. It is unbounded. And an unbounded model is never finished; it is only larger.
The First Model Scope Checklist (for Eclipse Capella) is the gate you pass before you open a new project. It is a workbook and a short guide that make you name the operational story, write down what is in and — just as importantly — what is out, tick only the Operational Analysis content this increment will actually show, and write a stop point that a named review can accept. Decide the boundary on paper. Then build.
Get it here: First Model Scope Checklist (for Eclipse Capella)
Why do first models stall?
Four problems account for most of it.
Everything is in scope. When nothing has been ruled out, everything is implicitly in. Every good question in a workshop becomes a new element, and every new element raises another good question.
Operational Analysis absorbs later layers. In the Arcadia method, Operational Analysis is the stakeholder story: entities and actors, capabilities, activities and interactions, processes, and scenarios — told before you specify what the system must do. The moment system functions, logical components, physical realization, or end items show up in the same sitting, the first increment has stopped being an operational story and started being an incomplete architecture.
The review has no stop point. A reviewer cannot accept “progress.” A reviewer can accept a defined increment against defined conditions. Without those, the review becomes a tour of the model and a list of things to add.
The boundary is discovered in the tool. Opening a fresh Capella project to find out what the increment includes is a natural instinct — the tool is right there, and it has so many helpful buttons. It is also the most reliable way to let the buttons decide the scope for you.
What does history say about scoped increments?
That the projects remembered for their discipline wrote the boundary down first — and the ones remembered for overruns often did not.
In June 1803, Thomas Jefferson gave Meriwether Lewis written instructions that began with the object of the mission: to explore the Missouri River and its principal tributary in search of the most direct and practicable water route across the continent. The instructions went on at length about what to observe and record, but the mission had a named purpose and a recognizable finish. The expedition’s journals are famous for their detail. Its scope was set before anyone stepped into a boat.
Apollo moved to the Moon in deliberately bounded steps. Apollo 8, in December 1968, orbited the Moon and came home without a lunar module. Apollo 10, in May 1969, took the full lunar landing configuration to within about nine miles of the surface — and did not land. Each mission had its own objectives and its own stop point, and a mission that met them was a success, even though “land on the Moon” was not yet in scope.
The Sydney Opera House is the counterexample everyone in project management eventually hears about. Construction began in 1959, before key parts of the design — the roof shells most famously — had been resolved. The building opened in 1973, ten years later than first hoped, and an early cost estimate of about A$7 million grew to roughly A$102 million. It is a magnificent building. It is also what happens when the boundary is discovered during construction.
A first Capella increment is a much smaller undertaking, but the logic scales down perfectly. Write the object. List what is out. Define what “done” means. Then build only that.
What may a first increment include?
The guide draws the line explicitly, and it is the most useful page in the kit.
In a first increment: a one-sentence operational story; named operational actors and entities; named operational capabilities; the activities and interactions that belong to those people and organizations; one process or one first scenario, if this increment includes it; and a named review with a stop point.
Out of a first increment: system functions and interfaces (System Analysis); logical components (Logical Architecture); physical realization, vendors, and screens (Physical Architecture and UI); end items, contracts, and IVVQ (EPBS); any OA-to-SA transition as part of Day 1; and anything you do not yet know — listed as out or unknown rather than left implicit.
One rule deserves special mention. The system under design gets a name on the Scope sheet, and then stays out of the operational story. It does not appear as a screen, a vendor, or a component. In this increment, the work is done by the people and organizations in the story. The system will appear in later layers, when those layers are in scope. Naming it, then deliberately setting it aside, is what keeps it from leaking in.
What is in the kit?
The zip contains a 4-page guide, a workbook, and a README.
**First-Model-Scope-Checklist.xlsx** has seven sheets:
- Start Here — the order to fill things in, and where the kit fits.
- Scope — increment name, the one-sentence operational story, who the increment is for, a real date or named review, the system under design as a name only, and one sentence for what “done for this increment” means.
- In and Out — a row for every kind of content, including every later Arcadia layer, so each one is explicitly placed in or out. The header says it plainly: if it is not on this sheet, it is not in the increment until you add it here first.
- Diagrams this increment — tick only the OA content the In list can support: capabilities; entities and actors and their involvements; activities and interactions; one process; one first scenario.
- Stop point — seven conditions, each with a Met? column and an Evidence column.
- Review — who will look, when, what they will be asked to accept, which questions are already known to be out of scope, and the follow-up increment if one is named.
- Example — Ferry Dock — a filled, fictional example.
The guide explains what a first increment is allowed to include, how to fill the workbook, the fictional Ferry Dock example as a narrative, and what to do once the stop conditions are met.
The kit does not include a .capella project file, and that is deliberate: a Capella project should be created and saved by Capella itself. You create it in Capella 7.1 after the increment is scoped. The kit also does not invent click-by-click menu paths for diagrams; you use Capella’s own Operational Analysis diagrams for the content you have already decided is in.
What are the seven stop conditions?
These are the finish line, and they are worth reading in full because each one closes a specific leak:
- The one-sentence operational story is written and unchanged for this increment.
- Every In item has a name, and every Out item is listed rather than left implicit.
- The system under design is named only, and has not leaked into the operational story as a screen, vendor, or component.
- The diagrams you ticked can be produced from the In list without adding new elements on the fly.
- A scenario or process, if one is in, traces to a named capability and to the entities or actors who perform the work.
- A named review will look at this increment against this checklist, and its date or event is on the Scope sheet.
- No OA-to-SA transition, and no later-layer diagrams, are part of this increment’s stop point.
Seven conditions, all checkable, none subjective. When they are all met, the increment is finished — not when later layers have been started.
What does the fictional Ferry Dock example show?
Northshore Ferry Dock is a made-up slip and operation, labelled fictional throughout, with no real operator, timetable, or performance claim. Its increment is a weather-hold announcement told as an operational story: when weather requires it, the dock crew announces a hold to passengers already on the dock, consistently and in time for people to stay clear of the ramp. The review audience is the dock-operations lead, on a fictional review date. The system under design is named once on Scope — a dock operations support system — and never drawn as screens or components.
The In list holds dock crew and passengers as actors, the weather-hold announcement capability, the announce-hold activity and the interaction that carries the hold to passengers, and one first scenario. The Out list is longer, and that is the point: system functions, a logical hold-management component, dock-side equipment, end items and contracts, ticketing, vessel scheduling, and any screen layout. Every tempting tangent is written down and parked.
You read the example last, after filling your own rows — so you write your own increment, not a lightly edited ferry dock.
What does the first hour look like?
Open the workbook on Start Here. Fill Scope. If you cannot write the “done for this increment” sentence, stop there: the increment is not yet scoped, and opening a Capella project will not help you find it. Fill In and Out, writing what is out rather than leaving rows blank. Tick only the diagrams the In list can support. Fill the Stop point and Review sheets. Read the Ferry Dock example, then revise your own rows if it sharpened anything.
When the stop conditions are met, create the project in Capella 7.1 and build only what the In list names. If an element is tempting simply because Capella has a button for it, check In and Out first.
How does it sit next to the other Capella tools?
The Operational Analysis Day-1 Project kit is the build: worksheets for actors, capabilities, activities, and one scenario, transcribed into Capella 7.1. This checklist is the gate you pass before that build — or the knife you use to cut back a model that has already grown. The Arcadia Layer Pocket Guide is the map of OA, System Analysis, Logical Architecture, Physical Architecture, and EPBS. Scope here, layers in the pocket guide, the OA build in the Day-1 kit. They work together, and none includes the others.
Capella is a trademark of Eclipse Foundation AISBL. Arcadia is the method Capella supports. This is an independent tool for people who use Eclipse Capella. It is not an official Eclipse or Capella product, and it does not include the Capella application or any official sample model.
A first model that can actually be finished
You are not paying for a sample model. You are paying for a workbook that forces an In/Out boundary with a row for every later layer, seven stop conditions a named review can check, a fictional example that shows what a disciplined first increment looks like, and a guide that keeps Day 1 on the Operational Analysis side of the line. It is a small, practical guard against the most common way first models stall.
Get the First Model Scope Checklist (for Eclipse Capella): First Model Scope Checklist (for Eclipse Capella)
It will be listed on our Tools for Eclipse Capella page alongside the rest of the tools in that line. You can browse every Organa Obscura tool on the Products page.
Leave a Reply