Chapter at a Glance

Engineering starts with an operational need and reaches evidence through explicit definitions.

Learn / Chapter 01

From Engineering Intent to Evidence

What would convince you that a monitoring configuration provides enough coverage?

Learning objective: distinguish an operational need, a measurable requirement, a model and the evidence from a particular execution.

Begin with the operator’s problem

An operator needs to understand activity within a bounded industrial area. Objects may enter, cross and leave it. Before choosing equipment, an engineering team must ask what the operator needs to observe and what would count as adequate monitoring. A sensor specification alone cannot answer those questions.

Our fictional Distributed Monitoring System starts with a smaller question: how much of a designated area lies within an assumed observation footprint? This reduction makes one engineering claim inspectable. It does not resolve the wider monitoring problem, which also concerns objects, events and decisions over time.

This is engineering intent: the outcome someone needs, before a particular solution is prescribed. The repository records it as NEED-001. Its identifier lets us refer back to the need; the identifier is not a substitute for understanding it. Choosing a single circular sensor is a modeling decision, not something the need itself demands.

Turn “enough” into an explicit criterion

“Provide coverage” leaves too much open for a reproducible evaluation. We need a defined area, a measure and a threshold. For this exercise, the proposed requirement is at least 50% geometric area coverage under a reference scenario and the stated V0 assumptions. Read its exact wording in REQ-COV-001.

The threshold is a teaching assumption. It was not derived from an operational analysis and is not a recommended design target. Making it explicit allows an experiment; it does not make the number authoritative. A real project would need a justified relationship between the criterion and the operator’s task.

The reference scenario, SCN-001, defines a 100 × 100 rectangle, a sensor at (50, 50) and a radius of 35 model units. Separating these inputs from the requirement lets us distinguish changing a candidate configuration from changing what it must satisfy.

Make a limited claim executable

The model treats the sensor footprint as a perfect disk and every part of the rectangle as equally important. Coverage means the fraction of the rectangle inside that disk. Distances use generic model units; no relationship to a physical sensor’s range has been established.

MODEL-COV-001 estimates that fraction with a fixed grid of 100 × 100 equal-area cells. It tests each cell’s center against the circular boundary, counts covered centers and converts that count to a percentage. Boundary points count as covered; a zero radius gives zero coverage. The smooth circle in the lab illustrates the ideal footprint, while the number is a sampled estimate with discretization error.

The model omits terrain, occlusion, directional sensitivity, uncertainty, time, faults and communication effects. These are limits on interpretation, not details to forget once a result appears. A coverage percentage is neither a probability of detecting an object nor a measure of tracking performance.

The reference configuration produces 38.52% estimated coverage, below the 50% criterion. Its footprint leaves too much of the rectangle uncovered. This intentional failure gives us a concrete discrepancy to investigate.

Experiment: one requirement, several configurations

Predict the effect of each change before running it. Keep the threshold and area fixed throughout.

  1. Record the reference configuration in the comparison table.
  2. Change only the radius until the criterion passes. Record that configuration.
  3. Keep that radius and move the sensor toward a corner. Record the changed result. Where does some of the footprint go?
  4. Find another passing configuration. Compare the inputs, percentage and margin, then explain which conclusion the model supports.

Each row is a separate trial of one sensor. The rows are not simultaneous sensors. Reset restores the reference inputs; Clear comparison removes your temporary notes.

One sensor. One explicit criterion.

Initialized from SCN-001. Change position or radius, then record separate trials below to compare them against the same criterion.

Interactive browser what-if run. These changes are temporary. They do not edit the source scenario or create saved CI verification evidence.
Ideal single-sensor coverageA rectangular monitoring area with one circular footprint, clipped to the area.

Area: 100 × 100 model units. Origin is the upper-left corner; x increases right and y increases down. Filled disk: ideal footprint. Dark dot: sensor position.

Sensor configuration (model units)

Use arrow keys or type whole model units. x: 0–100; y: 0–100; radius: 0–200. Position controls stay within the area; the engineering model also accepts outside positions.

FAIL — 38.52% estimated coverage; threshold 50.00%.

3852 / 10000 cell centers covered. Margin: -11.48 percentage points.

Compare separate single-sensor trials

Record up to four configurations. Each row is a separate one-sensor what-if run with the same area, threshold and model; rows do not combine sensors. Reset keeps your rows. Reloading or leaving this page discards them.

No configurations recorded.

Temporary browser comparisons — not retained CI evidence
TrialxyRadiusCoverage estimateMargin (pp)Criterion

Positions and radii use model units; margins use percentage points. A larger margin does not establish optimality: cost, installation constraints and real sensing performance are not evaluated.

Definition and interpretation

REQ-COV-001 defines the 50% criterion. This threshold is a teaching assumption, not an operationally derived target. TEST-COV-001 evaluates the reference scenario using MODEL-COV-001 v0.1.0.

Method: cell-center-grid, 100 × 100 equal-area cells. A cell is covered when its center’s distance is ≤ radius; zero radius means zero coverage. This is a sampled estimate, with discretization error. The smooth circle depicts the ideal footprint, not the sampled cells. PASS uses the unrounded estimate and includes equality.

  • One fixed ideal circular footprint in a uniformly weighted rectangular area.
  • No terrain, occlusion, uncertainty, time, faults, or communication effects.

No directional sensitivity or real sensor calibration is modeled. Area coverage is not detection probability, tracking performance or mission effectiveness. A PASS only satisfies this idealized geometric criterion.

Separate the result from the claim

The evaluator compares the unrounded estimate with the same threshold for every trial. Equality passes. A passing altered configuration shows that those inputs satisfy the geometric criterion; it does not replace the retained result for the unchanged reference scenario. Browser edits do not modify the canonical YAML.

Two configurations can both pass with different margins. That does not identify an optimum. We have defined no cost, power budget, installation constraint or operational objective to optimize. A larger assumed radius may improve this metric, but the model does not establish whether such a sensor exists or is suitable. Ask what additional criteria would be necessary before recommending either configuration.

Verification asks whether the specified criterion is met under the defined conditions. TEST-COV-001 executes the reference model and evaluates its requirement outcome independently of the website. Other automated tests compare known geometries and exercise invalid inputs and boundaries. A passing software test suite can correctly report a failing engineering configuration.

Evidence is the result of a particular execution together with enough context to interpret it. A retained reference record identifies the requirement, scenario, test, model version, inputs, sampling method, threshold and result. It also records the source commit, working-tree status, execution context and timestamp. Without those, an isolated PASS label cannot tell you what was evaluated.

To inspect retained evidence, follow the reference page’s CI artifact instructions. To reproduce it locally, use the repository’s test and evidence commands. Identical definitions and calculation outputs should repeat; timestamps and execution identifiers naturally differ. Temporary browser comparisons are exploratory notes, not retained CI artifacts.

Successful verification within this model is a narrow conclusion. Real-world suitability would require justified requirements and evidence about effects this model excludes. Keep the chain visible: intent motivates the criterion; definitions make it executable; evidence records what happened; interpretation limits the claim.

Check your reasoning

Answer before opening each explanation.

1. The software tests pass, but the reference requirement fails. Is that a contradiction?

No. The tests can confirm that the model and evaluator correctly report a configuration below the threshold. Software correctness and requirement satisfaction are different claims.

2. Two configurations pass. What would justify choosing one?

Additional, justified decision criteria and evidence: for example installation constraints, cost or actual sensing performance. Geometric PASS alone establishes neither optimality nor suitability.

3. Which is an assumption: the operator’s need, the ideal circular footprint, or a recorded coverage result?

The ideal circular footprint is a modeling assumption. The need expresses intent; the recorded percentage is an execution result. The threshold’s chosen value also has an explicitly assumed teaching basis.

4. What would you need to reproduce a saved result?

The source revision, model version, exact inputs, method and resolution, and evaluation threshold. The record’s timestamp and execution context identify the run; they are not numerical inputs to the deterministic calculation.

The next question: what must stay connected?

If the threshold changes, which evaluations need rerunning? If an assumption becomes invalid, which conclusions need review? Those questions require explicit relationships between engineering definitions and their evidence. Follow the source-to-test chain to inspect the small set we already maintain. Connected engineering knowledge begins with preserving those meanings and dependencies, before introducing a larger representation.

Next: Engineering Information Models →

Continue experimenting in the standalone lab →