You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
It would be very useful to implement a method that collision checks spline trajectories. The core piece of technology that we are missing is a function that takes a trajectory::Spline as input and returns a list of times such that every point on the trajectory is within some radius of the points obtained by evaluating the trajectory at those times. (That was a mouthful.)
We have something like this in the CheckRamp function included with Hauser's parabolic smoother. This function constructs a piecewise linear approximation of a parabolic spline, then runs a bisection collision check on the resulting piecewise linear trajectory. This is a reasonable approach that is described in Section IV.E. of the accompanying paper.
If we had function, then it would be relatively straightforward to replace significant parts of the current parabolic smoothing implementation, which is currently built as an external project, with the Aikido equivalents. It would also let us experiment with other spline-based trajectory optimization and retiming algorithms.
The text was updated successfully, but these errors were encountered:
It would be very useful to implement a method that collision checks spline trajectories. The core piece of technology that we are missing is a function that takes a
trajectory::Spline
as input and returns a list of times such that every point on the trajectory is within some radius of the points obtained by evaluating the trajectory at those times. (That was a mouthful.)We have something like this in the
CheckRamp
function included with Hauser's parabolic smoother. This function constructs a piecewise linear approximation of a parabolic spline, then runs a bisection collision check on the resulting piecewise linear trajectory. This is a reasonable approach that is described in Section IV.E. of the accompanying paper.If we had function, then it would be relatively straightforward to replace significant parts of the current parabolic smoothing implementation, which is currently built as an external project, with the Aikido equivalents. It would also let us experiment with other spline-based trajectory optimization and retiming algorithms.
The text was updated successfully, but these errors were encountered: