Main repo for developing traction adaptive motion planning using sampling augmented adaptive RTI. Under development!
The TAMP algorithm uses sampling augmented adaptive RTI to allow dynamically setting the tire force constraints. This enables the motion planner to deal with locally varying traction conditions in critical maneuvers. Details on the algorithm are available here: https://arxiv.org/abs/1903.04240. Simulations are done using https://github.com/AMZ-Driverless/fssim and the RTI solver is exported by the acado toolkit https://github.com/acado/acado.

System configuration: ubuntu 16.04 LTS & ROS Kinetic
http://wiki.ros.org/kinetic/Installation/Ubuntu

dependencies:
apt-get install ros-kinetic-jsk-rviz-plugins

clone this repo and fork of fssim to a new catkin workspace
git clone git@github.com:larsvens/tamp_ws.git
git clone git@github.com:larsvens/fssim.git

build with catkin build

To run the racing demo:
source devel/setup.bash
roslaunch common bringup_gotthard_FSG.launch
roslaunch common experiment.launch exp_config:="gotthard_racing_nonadapt_config.yaml"
saarti saarti_node.launch
roslaunch common ctrl_interface.launch

If you find the code useful in your own research, please consider citing

@article{svensson2019adaptive,
  title={Adaptive trajectory planning and optimization at limits of handling},
  author={Svensson, Lars and Bujarbaruah, Monimoy and Kapania, Nitin and T{\"o}rngren, Martin},
  journal={arXiv preprint arXiv:1903.04240},
  year={2019}
}