Risk Matrix and Evidence Strength

The final capstone is not a sales pitch for your mission.

It is a technical argument about whether the plan is defensible inside its stated boundary.

That means the team must do two things at once:

1. show why the mission concept is reasonable;
2. expose the evidence limits that could still invalidate it.

A polished presentation that hides uncertainty is weaker than a plain report that makes uncertainty inspectable.

Risk is not a scary-word list

A useful risk statement connects:

condition or hazard
        ↓
possible failure
        ↓
mission consequence
        ↓
control or design response
        ↓
verification evidence

Example:

Gust conditions exceed the planning margin → aircraft position/capture stability degrades → required imagery may become unusable or the operation may become unsafe → mission is delayed or redesigned → current weather evidence and the approved threshold are checked before launch.

Now the risk can drive a decision.

Build a bounded risk matrix

You can use simple qualitative categories if the reasoning is explicit.

Do not waste time arguing whether “possible” should be a 2 or a 3. The important part is whether the control and evidence actually match the failure.

Evidence has different strength

Not every item in your dossier proves the same kind of claim.

Strong direct evidence

Examples:

• a measured value with known method and units;
• a dated official source for a current rule or restriction;
• a processing report tied to the dataset being reviewed;
• a tested simulator or supplied scenario result;
• an inspectable image or route artifact.

Indirect evidence

Examples:

• manufacturer specifications used to estimate a capability;
• a prior test from a similar configuration;
• a model or calculation based on assumptions.

Assumption

Examples:

• future weather will remain inside limits;
• site conditions will match the supplied map;
• a specific configuration will behave as expected without current verification.

Assumptions are not forbidden. They just cannot be presented as direct evidence.

Worked example: “the route is safe”

A team writes:

Our route is safe because it stays inside the green area on our diagram.

That is a conclusion without a sufficient evidence chain.

A reviewer should ask:

• What does the green area represent?
• Who defined it?
• Does it include obstacle clearance, people, recovery access, and relevant authority boundaries?
• Is the site information current?
• What happens if conditions change during the mission?

A stronger claim is narrower:

In the supplied classroom site model, the proposed route remains inside the team-defined operating polygon and outside the mapped obstacle buffers. This supports the route-geometry portion of the plan only; current site conditions, people, weather, and applicable operational authorization would still require verification before any real flight.

That is a defendable claim because it says exactly what the evidence supports.

Residual risk matters

A control rarely makes risk disappear completely.

If the team adds a geofence, for example, residual concerns may include:

• incorrect boundary configuration;
• position error;
• home-point error;
• unexpected system behavior;
• a route that is technically inside the fence but still operationally poor.

The final dossier should show that you understand what remains after the control.

Build the final risk register

Choose at least six meaningful risks across different layers:

• authority/airspace;
• weather/site;
• people/crew;
• aircraft/energy;
• sensing/data;
• mission software/failsafe;
• product quality/evidence.

For each one, record:

• failure path;
• consequence;
• control;
• verification evidence;
• residual concern;
• owner or decision point.

Defense checkpoint

For every important claim in your mission plan, ask:

Is this supported by direct evidence, indirect evidence, or an assumption?

If the team cannot answer, the claim is not ready for defense.