Activity 6.18: Test and Iterate an Auto

🎯 Goal

By the end of this activity you will have:

• Selected an autonomous routine to iterate on
• Run the auto at least three times, logging data each run
• Identified and fixed at least one issue through observation and data analysis
• Documented the iteration cycle and improvements made

Before You Start

Complete Activity 6.17: Implement an Advanced Auto Feature. You need either a physical robot with a safe testing area or a working simulation environment. Have AdvantageScope ready for reviewing logged data.

---

Step 1: Choose Your Auto Routine

Pick an autonomous routine to iterate on. This can be:

• The auto you built in Activity 6.17
• An existing team auto that needs improvement
• A new auto you want to develop for competition

Auto routine I’m iterating on: _______________ Testing environment: ☐ Simulation ☐ Physical robot

---

Step 2: Prepare Your Logging

Before running the auto, make sure you’re logging the data you need to diagnose issues:

Essential Data to Log

// Add these to your drivetrain's periodic() if not already present
Logger.recordOutput("Auto/ActualPose", drivetrain.getPose());
Logger.recordOutput("Auto/TargetPose", pathFollower.getTargetPose());
Logger.recordOutput("Auto/PoseErrorMeters", poseError.getTranslation().getNorm());
Logger.recordOutput("Auto/HeadingErrorDeg", poseError.getRotation().getDegrees());
Logger.recordOutput("Auto/VelocityMps", chassisSpeeds.vxMetersPerSecond);

Pre-Run Checklist

Check	Done?
Logging is enabled and saving data
Auto routine is selected in the auto chooser
Robot starting position matches the path’s starting pose
All Named Commands are registered
Battery is charged (if on real robot)
Testing area is clear of obstacles and people

---

Step 3: Run #1 — Baseline

Run the auto routine and observe. Don’t try to fix anything yet — just collect data.

What to Watch For

• Does the robot start moving immediately or is there a delay?
• Does the robot follow the expected path shape?
• Does the robot reach each waypoint?
• Do mechanism actions (intake, shooter) fire at the right times?
• Does the robot end in the expected position and orientation?

Run #1 Log

Observation	What Happened
Starting behavior
Path following accuracy
Mechanism timing
End position accuracy
Total time
Issues noticed

---

Step 4: Analyze Run #1 Data

Open the log file in AdvantageScope and examine:

1. Pose tracking — overlay the actual pose on the target pose. Where does the robot deviate?
2. Velocity profile — is the robot reaching the expected speeds? Are there sudden changes?
3. Heading error — does the robot maintain the correct orientation throughout?
4. Event timing — do mechanism actions start at the right path positions?

Common Issues and Adjustments

Issue	What You See in Data	Adjustment
Robot overshoots turns	Large heading error after turns	Reduce max angular velocity in constraints
Robot doesn’t reach waypoints	Pose error grows over time	Check odometry calibration, tune translation PID
Mechanism fires too early/late	Event marker timing off	Move the event marker position on the path
Robot is too slow	Total time exceeds target	Increase max velocity (if safe)
Robot oscillates on path	Pose error oscillates around zero	Reduce translation PID P gain
Robot drifts sideways	Consistent lateral offset	Check gyro calibration, wheel alignment

---

Step 5: Make Adjustments

Based on your analysis, make ONE change at a time. This is critical — if you change multiple things, you won’t know which change fixed (or broke) something.

Adjustment I’m making: _______________ Why: _______________ What I expect to change: _______________

Types of Adjustments

Category	Examples
Waypoint positions	Move a waypoint to give the robot more room for a turn
Constraints	Reduce max velocity near scoring, increase on straightaways
Event marker timing	Move a marker earlier so the intake deploys before arrival
PID tuning	Adjust path following PID gains for tighter tracking
Starting position	Adjust the auto’s starting pose to match where the robot actually sits

---

Step 6: Run #2 — After First Adjustment

Run the auto again with your change applied.

Run #2 Log

Observation	What Happened	Better/Worse/Same?
Starting behavior
Path following accuracy
Mechanism timing
End position accuracy
Total time
Issues noticed

Did the adjustment help? _______________ Next adjustment needed: _______________

---

Step 7: Run #3 — After Second Adjustment

Make your second adjustment and run again.

Run #3 Log

Observation	What Happened	Better/Worse/Same?
Starting behavior
Path following accuracy
Mechanism timing
End position accuracy
Total time

---

Step 8: Compare Runs

Fill in the summary comparison:

Metric	Run #1	Run #2	Run #3
End position error (meters)
End heading error (degrees)
Total time (seconds)
Mechanism actions successful?
Overall quality (1–5)

▶Iteration Tips from Experienced Teams

Start slow, then speed up. Run the auto at 50% speed first to verify the path shape is correct. Then gradually increase to full speed.

Log everything. You can always ignore extra data, but you can’t go back and log something you didn’t capture.

One change at a time. It’s tempting to fix three things at once, but disciplined iteration is faster in the long run.

Save your logs. Name log files with the run number and what changed. You’ll want to compare them later.

Battery matters. On a real robot, a low battery changes motor behavior. Always test with a charged battery and note the voltage.

---

✅Checkpoint: Auto Iteration

Describe your iteration process: (1) What was the biggest issue you found in Run #1? (2) What adjustment fixed it? (3) How much did the auto improve from Run #1 to Run #3? (4) What would you do differently next time you iterate on an auto?

▶Check Answer

Strong answers include:

1. Specific issue — e.g., “The robot was overshooting the turn near the scoring position by about 0.3 meters, causing it to miss the target.”

2. Targeted fix — e.g., “I added a constraint zone on the last 25% of the path with max velocity reduced from 3.5 to 1.5 m/s. This gave the robot time to decelerate before the turn.”

3. Measurable improvement — e.g., “End position error went from 0.4m to 0.08m. Total time increased by 0.3 seconds due to the slower approach, but the accuracy improvement is worth it.”

4. Process learning — e.g., “Next time I’d start by running at 50% speed to verify the path shape before testing at full speed. I wasted Run #1 debugging a path shape issue that would have been obvious at low speed.”

---

📖 Key Vocab Used in This Activity

• Path Constraints — limits on velocity, acceleration, and other parameters that control how fast the robot drives a path
• Constraint Zone — a section of a path where different constraints apply (e.g., slowing down near a scoring position)
• Auto Routine — a complete autonomous sequence combining paths, Named Commands, and event markers

What’s Next?

You’ve practiced the full auto iteration cycle — the same process teams use at competition to refine their autonomous routines between matches. In Lesson 6.19: Competition Readiness, you’ll learn the complete competition preparation toolkit — pre-match checklists, pit debugging flowcharts, match log analysis, and hotfix branching strategies.