Ultra-Fast Trajectory Generation (Motion Profiling)

[Jaci]

Hi there,

A while ago, I wrote the Pathfinder (https://github.com/JacisNonsense/Pathfinder) library for Motion Profiling and Trajectory Generation. A little more recent, I wrote this post (http://imjac.in/ta/post/2016/11/14/08-00-15-generated/) on using a special part of the RoboRIO's ARM Architecture to do vision tracking more efficiently. Today, I'm combining the two and showing you how to generate Trapezoidal, S-Curve, both 1D and 2D Trajectories extremely fast.

Here (http://imjac.in/ta/post/2016/12/29/11-07-56-generated/) is the post, where you will find a discussion in a bit more detail, as well as the results obtained.

In short, we can generate profiles and trajectories very, very fast, a huge increase in speed over what is already close-to-metal C.

The code used in this investigation is available here (https://github.com/JacisNonsense/neon_pathfinder)

[marshall]

Hi there,

A while ago, I wrote the Pathfinder (https://github.com/JacisNonsense/Pathfinder) library for Motion Profiling and Trajectory Generation. A little more recent, I wrote this post (http://imjac.in/ta/post/2016/11/14/08-00-15-generated/) on using a special part of the RoboRIO's ARM Architecture to do vision tracking more efficiently. Today, I'm combining the two and showing you how to generate Trapezoidal, S-Curve, both 1D and 2D Trajectories extremely fast.

Here (http://imjac.in/ta/post/2016/12/29/11-07-56-generated/) is the post, where you will find a discussion in a bit more detail, as well as the results obtained.

In short, we can generate profiles and trajectories very, very fast, a huge increase in speed over what is already close-to-metal C.

The code used in this investigation is available here (https://github.com/JacisNonsense/neon_pathfinder)

Nifty stuff. Good work! I'm curious to see how this gets adapted for use by the community down the road. Seems like the basis for creating some new libraries and executables.