launches the following:
- Force torque sensor node, published wrench (ros package net-ft-ros)
- Python FT classifier, published feature labels (ros package ft_sensor_python)
- Virtual peg_sensor model, used to estimate expected measurements (not necessary when using FT) (ros package peg_sensor)
- Particle filter (ros package peg_filter)
The peg filter node does two things; 1) update the particle fitler and 2) compresses the particle filter to a lower dimensional representation. It published the particle filter weights which are shown in Rviz and the compressed feautres as well.
topics: [name,type]
- [mode_feature, std_msgs::Float64MultiArray>]
- [bel_x,visualization_msgs::Marker]
- [pfilter,sensor_msgs::PointCloud]
Initialised and launches the server actions. One action, plug_search, is added to the action server.
- call plug_search
- startes the plug_search action in the action server
- call gmm
- starts the GMM policy via a call to a service in the plug_search action.
- actions (velocity) to the robot controller
topics: [name, type]
- [ft_classifier,std_msgs::Float32MultiArray]
- virtual sensor topic: /virtual_classifier : [std_msgs/Float64MultiArray], distance to features
- ft sensor topic: /ft_classifier : [std_msgs/Float64MultiArray]
- rosrun rqt_reconfigure rqt_reconfigure