V2 by Superbro525Alt · Pull Request #39 · CurtinFRC/Repulsor

Superbro525Alt · 2026-02-17T12:44:07Z

No description provided.

Resolves #39

Signed-off-by: Jade Turner <spacey-sooty@proton.me>

Copilot

Pull request overview

Adds new “static shot” solving + offload entrypoints, and introduces a new Shuttle Recovery mode that uses predictive selection of recovery points and an offloaded FieldTracker endpoint.

Changes:

Expose a new DragShotPlanner static-position solve API (local/async/offloaded) and add the corresponding offload task ID + entrypoint.
Add shuttle recovery point selection (DTOs + predictive local/offloaded entrypoints) and a new FieldTracker API to request the next shuttle recovery goal.
Introduce SHUTTLE_RECOVERY_MODE with a new ShuttleRecoveryBehaviour, and update reasoning/behaviour selection wiring.

Reviewed changes

Copilot reviewed 22 out of 22 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
src/test/java/org/curtinfrc/frc2026/util/Repulsor/Offload/ShuttleRecoveryOffloadApiTest.java	Reflection-based API checks for new shuttle recovery offload wrappers/task IDs.
src/test/java/org/curtinfrc/frc2026/util/Repulsor/Offload/DragShotStaticPositionApiTest.java	Reflection-based API checks for new static-shot solve API and generated offload wrapper methods.
src/main/java/org/curtinfrc/frc2026/util/Repulsor/Tracking/FieldTrackerLocalAccess.java	Local fallback implementation for shuttle recovery goal computation (via predictive runtime).
src/main/java/org/curtinfrc/frc2026/util/Repulsor/Tracking/FieldTrackerCore.java	Adds recovery-dynamics snapshot + new offloaded `nextAllianceShuttleRecoveryGoalBlue` API.
src/main/java/org/curtinfrc/frc2026/util/Repulsor/Shooting/DragShotPlannerStaticPositionSearch.java	New static-position shot solve (coarse + refinement) implementation.
src/main/java/org/curtinfrc/frc2026/util/Repulsor/Shooting/DragShotPlannerLocalAccess.java	Adds local-access wrapper for static-position solve.
src/main/java/org/curtinfrc/frc2026/util/Repulsor/Shooting/DragShotPlannerCore.java	Adds core-level static-position solve entrypoint.
src/main/java/org/curtinfrc/frc2026/util/Repulsor/Shooting/DragShotPlanner.java	Public API for static-position solve (offloaded/async/local).
src/main/java/org/curtinfrc/frc2026/util/Repulsor/Reasoning/Rebuilt2026Reasoner.java	Adds shuttle recovery signals/phases and changes behaviour selection logic.
src/main/java/org/curtinfrc/frc2026/util/Repulsor/Predictive/PredictiveFieldStateRuntime.java	Adds local/offloaded shuttle recovery point selection APIs.
src/main/java/org/curtinfrc/frc2026/util/Repulsor/Predictive/PredictiveFieldStateLocalAccess.java	Implements local shuttle recovery point selection using predictive ranking.
src/main/java/org/curtinfrc/frc2026/util/Repulsor/Offload/ShuttleRecoveryPointDTO.java	DTO for shuttle recovery point selection result.
src/main/java/org/curtinfrc/frc2026/util/Repulsor/Offload/ShuttleRecoveryDynamicObjectDTO.java	DTO for dynamic-object snapshots used by shuttle recovery selection.
src/main/java/org/curtinfrc/frc2026/util/Repulsor/Offload/PredictiveFieldStateOffloadEntrypoints.java	Offloadable entrypoint for selecting shuttle recovery point.
src/main/java/org/curtinfrc/frc2026/util/Repulsor/Offload/OffloadTaskIds.java	Adds new task IDs for static shot + shuttle recovery selection and goal.
src/main/java/org/curtinfrc/frc2026/util/Repulsor/Offload/FieldTrackerOffloadEntrypoints.java	Offloadable entrypoint for shuttle recovery goal selection.
src/main/java/org/curtinfrc/frc2026/util/Repulsor/Offload/DragShotPlannerOffloadEntrypoints.java	Adds offloadable entrypoint for static-position shot solve.
src/main/java/org/curtinfrc/frc2026/util/Repulsor/Behaviours/ShuttleRecoveryBehaviour.java	New behaviour for shuttle recovery mode (collect + shoot routing and shot solve integration).
src/main/java/org/curtinfrc/frc2026/util/Repulsor/Behaviours/ShuttleBehaviour.java	Refactor/updates to shuttle mode behaviour, including shot solve integration.
src/main/java/org/curtinfrc/frc2026/util/Repulsor/Behaviours/BehaviourFlag.java	Adds `SHUTTLE_RECOVERY_MODE`.
src/main/java/org/curtinfrc/frc2026/util/Repulsor/Behaviours/AutoPathBehaviour.java	Tightens `shouldRun` to require `AUTOPATH_MODE`.
src/main/java/org/curtinfrc/frc2026/Robot.java	Wires in `ShuttleRecoveryBehaviour` and disables reasoner “testing” by default.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-02-17T12:49:40Z

+    EnumSet<BehaviourFlag> out =
+        hubActive
+            ? EnumSet.of(BehaviourFlag.SHUTTLE_RECOVERY_MODE)
+            : EnumSet.of(BehaviourFlag.SHUTTLE_MODE);


update() no longer calls seq.update(ctx) and instead always returns either SHUTTLE_MODE or SHUTTLE_RECOVERY_MODE (except TESTING). This makes the configured phases/transitions (autopath/defense/endgame + NT-controlled transitions) effectively dead code and prevents those modes from ever being selected via the reasoner. Consider restoring EnumSet<BehaviourFlag> out = seq.update(ctx) after updating the signals (e.g., set WANT_SHUTTLE/WANT_SHUTTLE_RECOVERY from hubActive but still let the SequenceReasoner choose the phase), or remove the sequence machinery entirely if it’s intentionally unused.

Suggested change

EnumSet<BehaviourFlag> out =

hubActive

? EnumSet.of(BehaviourFlag.SHUTTLE_RECOVERY_MODE)

: EnumSet.of(BehaviourFlag.SHUTTLE_MODE);

EnumSet<BehaviourFlag> out = seq.update(ctx);

Copilot · 2026-02-17T12:49:41Z

+public final class ShuttleRecoveryBehaviour extends Behaviour {
+  private static final double SHOOT_BEHIND_HUB_METERS = 2.95;
+  private static final double SHOOT_LATERAL_STEP_METERS = 0.45;
+  private static final double SHOOT_POS_TOL_METERS = 0.34;
+  private static final double SHOOT_YAW_TOL_DEG = 13.0;
+  private static final double FIELD_MARGIN_METERS = 0.28;
+
+  private static final double MOTION_COMP_LATENCY_SEC = 0.08;
+  private static final double MOTION_COMP_MIN_LEAD_SEC = 0.10;
+  private static final double MOTION_COMP_MAX_LEAD_SEC = 0.45;
+  private static final double MOTION_COMP_MAX_SPEED_MPS = 4.5;
+  private static final double DEFAULT_TIME_TO_PLANE_SEC = 0.18;
+  private static final long MAGAZINE_CAPACITY = 16L;
+  private static final int RECOVERY_GOAL_UNITS = 2;
+
+  private static final double[] SHOOT_LATERAL_OFFSETS =
+      new double[] {0.0, SHOOT_LATERAL_STEP_METERS, -SHOOT_LATERAL_STEP_METERS, 0.9, -0.9};
+
+  private static final GamePiecePhysics SHUTTLE_GAME_PIECE = loadShuttleGamePiece();
+
+  private final int prio;
+  private final Supplier<Boolean> hasPiece;
+  private final Supplier<Double> ourSpeedCap;
+
+  private final NetworkTablesValue<Double> shotAngle =
+      NetworkTablesValue.ofDouble(
+          NetworkTableInstance.getDefault(), NetworkTablesValue.toAdvantageKit("/ShotAngle"), 0.0);
+  private final NetworkTablesValue<Double> shotSpeed =
+      NetworkTablesValue.ofDouble(
+          NetworkTableInstance.getDefault(), NetworkTablesValue.toAdvantageKit("/ShotSpeed"), 0.0);
+  private final NetworkTablesValue<Boolean> shooterPassthrough =
+      NetworkTablesValue.ofBoolean(
+          NetworkTableInstance.getDefault(),
+          NetworkTablesValue.toAdvantageKit("/ShooterPassthrough"),
+          false);
+  private final NetworkTablesValue<Long> pieceCount =
+      NetworkTablesValue.ofInteger(NetworkTableInstance.getDefault(), "/PieceCount", 0L);


ShuttleRecoveryBehaviour duplicates a large amount of logic/constants/helper methods from ShuttleBehaviour (shot solving, motion compensation, NetworkTables outputs, clamp helpers, etc.). This duplication will be hard to keep consistent as tuning changes land. Consider extracting the shared aiming/shot-selection code into a shared helper (or a base class) that both behaviours can call, leaving only the collect-goal selection differences in each behaviour.

Copilot · 2026-02-17T12:49:41Z

+  static Optional<ShotSolution> calculateShotAngleAndSpeedFromStaticPosition(
+      GamePiecePhysics gamePiece,
+      Translation2d shooterFieldPosition,
+      Translation2d targetFieldPosition,
+      double targetHeightMeters,
+      double shooterReleaseHeightMeters,
+      Constraints constraints) {
+
+    AutoCloseable _p = Profiler.section("DragShotPlanner.calculateStaticShotAngleAndSpeed");
+    try {
+      if (constraints == null) {
+        Profiler.counterAdd("DragShotPlanner.static.constraints_null", 1);
+        return Optional.empty();
+      }
+
+      double minSpeed = constraints.minLaunchSpeedMetersPerSecond();
+      double maxSpeed = constraints.maxLaunchSpeedMetersPerSecond();
+      double minAngleDeg = constraints.minLaunchAngleDeg();
+      double maxAngleDeg = constraints.maxLaunchAngleDeg();
+      Constraints.ShotStyle shotStyle = constraints.shotStyle();
+      boolean fixedAngle = Math.abs(maxAngleDeg - minAngleDeg) < 1e-6;
+
+      if (minSpeed <= 0.0 || maxSpeed <= minSpeed) {
+        Profiler.counterAdd("DragShotPlanner.static.bad_speed_range", 1);
+        return Optional.empty();
+      }
+      if (!fixedAngle && maxAngleDeg <= minAngleDeg) {
+        Profiler.counterAdd("DragShotPlanner.static.bad_angle_range", 1);
+        return Optional.empty();
+      }
+
+      double acceptableError = DragShotPlannerConstants.ACCEPTABLE_VERTICAL_ERROR_METERS;
+      ShotSolution coarse =
+          DragShotPlannerSolveAtPosition.solveBestAtShooterPosition(
+              gamePiece,
+              shooterFieldPosition,
+              targetFieldPosition,
+              targetHeightMeters,
+              shooterReleaseHeightMeters,
+              minSpeed,
+              maxSpeed,
+              minAngleDeg,
+              maxAngleDeg,
+              fixedAngle,
+              acceptableError,
+              shotStyle,
+              0.35,
+              0.6);
+      if (coarse == null) {
+        return Optional.empty();
+      }
+
+      double coarseErrorAbs = Math.abs(coarse.verticalErrorMeters());
+      if (coarseErrorAbs <= acceptableError) {
+        return Optional.of(coarse);
+      }
+


The new static-position solver (calculateShotAngleAndSpeedFromStaticPosition) adds non-trivial selection/refinement logic, but the tests added in this PR only assert API/wrapper presence via reflection. Adding at least one unit test that exercises the solver (e.g., known shooter/target geometry + constraints) would help catch regressions and validate that the coarse/refine steps behave as intended.

TunerConstants already includes an instance of the Drive CANBus, using it is a bit clearer and should reduce memory usage ever so slightly. Signed-off-by: Jade Turner <spacey-sooty@proton.me>

We had these enabled last year. Signed-off-by: Jade Turner <spacey-sooty@proton.me>

Signed-off-by: Jade Turner <spacey-sooty@proton.me>

* feature: add ball sim Co-authored-by: Jade <spacey-sooty@proton.me> * feature: implement ball sim in shooter sim --------- Co-authored-by: Jade <spacey-sooty@proton.me>

Co-authored-by: Jade Turner <spacey-sooty@proton.me>

Co-authored-by: Jason <jy201016@gmail.com>

* Created MagIOComp layer, added a second constructor to Mag.java to account for B-Bot, along with if-else statements to change between those constructors in the Commands, and integrated it into Robot.java * Fixed one formatting issue

* Added IntakeIOComp layer and integrated it into Robot.java * Pulled from main * Built code

Signed-off-by: Jade Turner <spacey-sooty@proton.me>

…ous hub facing based on location (#85) * Added IntakeIOComp layer and integrated it into Robot.java * Pulled from main * Built code * Attempted to create a trigger to only allow shooting when facing the target location. Not finished * Not finished * Autonomous facing hub when in Alliance Zone * Added new tuning file, tested driving, and added triggers for autonomous hub facing

Signed-off-by: Jade Turner <spacey-sooty@proton.me>

* chore: do review comments * fix: update hood gear ratios * fix: remove sysid and add recalc hood feedforward values * fix: add recalc gains for hooded shooter * chore: move magic numbers into constants * feature: pid gains and general values for hoodedshooter * fix: hood pid --------- Co-authored-by: prawny-boy <seanyk.chan@gmail.com>

Eason-F and others added 2 commits February 9, 2026 17:05

feature: add FieldConstants from MechanicalAdvantage

c0c5b31

Resolves #39

chore: update all dependencies to latest (#57)

b160038

Signed-off-by: Jade Turner <spacey-sooty@proton.me>

Copilot AI review requested due to automatic review settings February 17, 2026 12:44

Copilot started reviewing on behalf of Superbro525Alt February 17, 2026 12:44 View session

Copilot AI reviewed Feb 17, 2026

View reviewed changes

spacey-sooty and others added 9 commits February 26, 2026 18:58

fix: Use TunerConstants CANBus (#64)

67d67b8

TunerConstants already includes an instance of the Drive CANBus, using it is a bit clearer and should reduce memory usage ever so slightly. Signed-off-by: Jade Turner <spacey-sooty@proton.me>

fix: enable rio2 memory settings (#62)

afef65a

We had these enabled last year. Signed-off-by: Jade Turner <spacey-sooty@proton.me>

fix: Fix command logging (#63)

56254a6

Signed-off-by: Jade Turner <spacey-sooty@proton.me>

feature: Add LoggedNetworkStruct<T> (#70)

d8c65a5

Signed-off-by: Jade Turner <spacey-sooty@proton.me>

feature: add ball sim (#66)

a0583db

* feature: add ball sim Co-authored-by: Jade <spacey-sooty@proton.me> * feature: implement ball sim in shooter sim --------- Co-authored-by: Jade <spacey-sooty@proton.me>

feature: add LoggedTunableNumber for tuning in advantagescope (#67)

747ea87

Co-authored-by: Jade Turner <spacey-sooty@proton.me>

feature: face hub and hub heading calculations for drive.java (#68)

2ba12b7

Co-authored-by: Jason <jy201016@gmail.com>

fix: add apriltag to fieldconstants path (#71)

9d27af4

fix: Decrease camera FPS limit when disabled (#73)

0241178

Superbro525Alt changed the title ~~Extend shot~~ V2 Feb 27, 2026

Eason-F and others added 15 commits February 27, 2026 20:21

feature: add comp io layers (#69)

3614491

feature: added game state tracking for the hub state (#56)

1edef1d

Added IntakeIOComp layer and integrated it into Robot.java (#75)

037f1e8

* Added IntakeIOComp layer and integrated it into Robot.java * Pulled from main * Built code

chore: update ports for beta bot (#79)

29b70a3

Signed-off-by: Jade Turner <spacey-sooty@proton.me>

chore: update TunerConstants for comp bot (#80)

c445f78

Signed-off-by: Jade Turner <spacey-sooty@proton.me>

fix: Initialise HoodedShooter properly (#81)

5cbd81f

Signed-off-by: Jade Turner <spacey-sooty@proton.me>

Push (#82)

6813828

chore: enable Phoenix and DataLogManager loggers (#84)

9d92ddd

Signed-off-by: Jade Turner <spacey-sooty@proton.me>

fix: Make alliance empty report correctly (#83)

54bd829

Signed-off-by: Jade Turner <spacey-sooty@proton.me>

update

a2e4907

W speed

945b0a2

idk bro

27385bc

Superbro525Alt added 27 commits March 1, 2026 15:26

remove calibration things

df83523

a

9d9c8d0

a

0edebb4

a

182f551

a

74022bf

a

41a506e

a

cf9f137

a

e5e900c

a

4aebccb

a

b312b9b

a

d35b6e4

a

b3fc6ee

a

99e0f9a

a

22021b2

a

8a64e59

a

112dda6

Fix ci

fb64f56

ready to test

818f59b

a

979e787

a

39a4dfd

a

5b48be7

update deps

e945e31

a

41d2f62

a

5c73df7

a

deb1fb5

a

d943239

merge

28c140a

Superbro525Alt force-pushed the extend-shot branch from ef33086 to 28c140a Compare March 1, 2026 08:35

a

5fd3212

Superbro525Alt closed this Mar 1, 2026

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

V2#39

V2#39
Superbro525Alt wants to merge 204 commits intomainfrom
extend-shot

Superbro525Alt commented Feb 17, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Feb 17, 2026

Uh oh!

Copilot AI Feb 17, 2026

Uh oh!

Copilot AI Feb 17, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

Conversation

Superbro525Alt commented Feb 17, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Feb 17, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 17, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 17, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants