Path Constraints

The limits in your drivetrain constants describe what the robot can do everywhere. Path constraints let you lower those limits for a particular part of one path.

Think of them like speed-limit signs. The robot may normally be capable of moving at 60 inches per second, but a constraint can limit it to 30 inches per second before a narrow or precise section. The lower limit remains active until another constraint of the same type replaces it.

Where a Constraint Starts

Every constraint has an s value from 0.0 to 1.0 describing its position along the path. An s of 0.0 is the start, 0.5 is halfway by distance, and 1.0 is the end. Once the robot passes that point, the constraint becomes active.

For example, this lowers the translational velocity limit after the robot reaches 25% of the path:

.addConstraint(new TranslationalConstraint(
        0.25,
        ConstraintType.VELOCITY,
        Dist.fromIn(30)
))

To restore a higher limit later, add another velocity constraint:

.addConstraint(new TranslationalConstraint(
        0.75,
        ConstraintType.VELOCITY,
        Dist.fromIn(55)
))

Constraint Types

A TranslationalConstraint limits movement along the path. With ConstraintType.VELOCITY, it limits linear speed. With ConstraintType.ACCELERATION, it limits how quickly that speed may change.

An AngularConstraint applies the same idea to heading. Its velocity form limits how quickly the robot may rotate, while its acceleration form limits how quickly that rotation rate may change.

.addConstraint(new TranslationalConstraint(
        0.20, ConstraintType.ACCELERATION, Dist.fromIn(40)
))
.addConstraint(new AngularConstraint(
        0.20, ConstraintType.VELOCITY, Angle.fromRad(2.0)
))

A Profiled Build uses all four forms while generating its velocity and acceleration plan. A holonomic Quick Build only uses translational velocity constraints because it does not generate the complete profile needed to enforce acceleration and angular limits ahead of time.

useBoostedAccel()

A normal motion profile assumes the robot begins and ends the path at zero velocity. This gives the generator enough room to plan a controlled acceleration away from the start and a controlled slowdown before the end.

useBoostedAccel() removes those zero-velocity boundary conditions. The profile may begin and end at its allowed cruising speed, letting velocity feedback request much more acceleration at the start and much more braking near the end. The follower still uses the translational PDS controller applied tangentially to the path to prevent the robot from driving through the final point.

This does not increase the configured velocity or acceleration limits. Instead, it stops the profile from reserving distance for a gradual start and stop, allowing the feedback controller to use the available power more aggressively.

Boosted acceleration is useful when completing a movement as quickly as possible matters more than a gentle transition. It is generally recommended only for robots with robust localization such as a two wheel odometry system to useBoostedAccel() because the wheel slip involved in aggressive movements prevents drive encoders from reporting accurate positional data.

To use the boosted acceleration, simply call useBoostedAccel() anywhere in the path builder API:

Path examplePath = new Builder.holonomicPath(
    pose.of(0,0),
    pose.of(10,30))
    .useBoostedAccel()
    .profiledBuild();