Small Fixes

- Add "/" to front of topic name in cv_model
- Change default spawn point to be closer to depth camera
- Fix desired grasp pose offset

src/rrrobot_ws/src/rrrobot/config/rrrobot_sensors.yaml

@@ -1,9 +1,4 @@
 sensors:
-    logical_camera_1:
-      type: logical_camera
-      pose:
-        xyz: [1.2, 3.5, 1.5]
-        rpy: [0, 1.5707, 0]
     depth_camera_1:
       type: depth_camera
       pose:

src/rrrobot_ws/src/rrrobot/src/cv_model.py

@@ -50,7 +50,7 @@ def call_back(filename):
     print('type: ', type_dict[predicted_label])
 
     # publish a message, name of this node is 'cv_model'
-    pub = rospy.Publisher('cv_model', String, queue_size=10)
+    pub = rospy.Publisher('/cv_model', String, queue_size=10)
     # rospy.init_node('talker', anonymous=True)
     rate = rospy.Rate(10) # 10hz
 

src/rrrobot_ws/src/rrrobot/src/object_spawner_node.cpp

@@ -20,7 +20,7 @@ class ObjectSpawner
 {
 public:
     ObjectSpawner(const std::string &model_file = "/app/rrrobot_ws/src/gazebo_models/model_mappings.txt")
-        : DEFAULT_SPAWN_POINT(1.21825, 5.474367, 0.937978),
+        : DEFAULT_SPAWN_POINT(1.21825, 4.0, 0.937978),
           CONVEYOR_WIDTH(0.391404)
     {
         // DEFAULT_SPAWN_POINT.x = 0.2516105;

src/rrrobot_ws/src/rrrobot/src/rrrobot_node.cpp

@@ -127,7 +127,7 @@ public:
 
 		desired_grasp_pose = grasp_pose;
 		// TODO: Tune z offset so end effector doesn't hit object
-		desired_grasp_pose.position.z -= 0.025;
+		desired_grasp_pose.position.z += 0.01;
 
 		if (current_robot_state & RobotState::WAITING_FOR_CLASSIFICATION)
 		{