Reproduce a flight.
Choose a recorded experiment. Keep its revision, seeds and settings fixed. Report agreements and differences, including your environment.
OPEN QUESTIONS / CONTRIBUTIONS
Bring a better question.
Useful contributions can be small. An independently reproduced failure can teach more than another controller feature.
Challenge an assumption.
Bring vehicle-response data, a communication model, a geometry correction or a reason a measurement does not answer its question.
Design a difficult encounter.
Describe a scenario, the mechanism you expect to fail and the observable result that would change your mind.
Make the evidence easier to see.
Improve a chart, explain a record or help distinguish contact, damage, loss and delivery performance.
Start with a research issue.
Include the question, relevant flight-log entry, source revision, expected result and a bounded proposed test. If you ran it already, attach the result and describe any deviations from the plan.
Repository access and contribution permissions depend on its publication and license status. See the repository’s current terms before redistributing code.
Preserve negative results and corrections. Keep a prediction separate from a result, and label exploratory analysis when it was chosen after seeing the data.