Vision Fallbacks and Confidence

Make autonomous robust when vision is uncertain.

Module 11: Vision and AprilTagsVision

In this lesson, you will:

01Define confidence.
02Choose default behavior.
03Log selected branch.

Concept narrative

Vision should not be all-or-nothing. The robot needs a plan for low confidence, no detections, or conflicting readings.

Robot mental model

A safe fallback is part of the vision system. It tells the robot what to do when the camera cannot provide a trustworthy answer.

Implementation walkthrough

Add confidence or age of detection to the processor result. In autonomous, choose a default branch and print the selected reason.

VisionResult.javaJava

PropPosition result = processor.getConfidence() > 0.65
    ? processor.getPosition()
    : PropPosition.CENTER;
telemetry.addData("auto branch", result);

Common mistakes and debugging

If the default is unsafe, the robot will fail exactly when uncertainty is highest. Make fallback conservative and tested.

Checkpoint

Unknown result has a default.
Branch reason is logged.
Fallback is tested.
The test record includes the setup, prediction, and observed result.
A teammate can repeat the check from the saved evidence without guessing.

Reflection check

Check your understanding before moving on.

Which result best demonstrates completion of “Vision Fallbacks and Confidence”?

Why record the test setup, prediction, and observed result?

0 of 2 answered

References

FTC VisionPortal DocsOfficial FTC camera and processor lifecycle documentation.GM0 Computer VisionFTC-oriented vision concepts, pipelines, and practical advice.FTC AprilTag DocsOfficial AprilTag detection and metadata reference.