diff --git a/contrib/ros/envs/example_launchers/trpo_gazebo_sawyer_pnp.py b/contrib/ros/envs/example_launchers/trpo_gazebo_sawyer_pnp.py
index 0f269077c..1f2b6f202 100644
--- a/contrib/ros/envs/example_launchers/trpo_gazebo_sawyer_pnp.py
+++ b/contrib/ros/envs/example_launchers/trpo_gazebo_sawyer_pnp.py
@@ -1,3 +1,5 @@
+import rospy
+
 from rllab.baselines.linear_feature_baseline import LinearFeatureBaseline
 from rllab.envs.normalized_env import normalize
 from rllab.misc.instrument import run_experiment_lite
@@ -10,6 +12,8 @@
 
 
 def run_task(*_):
+    rospy.init_node('trpo_sim_sawyer_pnp_exp', anonymous=True)
+
     env = TfEnv(normalize(PickAndPlaceEnv()))
 
     policy = GaussianMLPPolicy(
diff --git a/contrib/ros/envs/example_launchers/trpo_gazebo_sawyer_push.py b/contrib/ros/envs/example_launchers/trpo_gazebo_sawyer_push.py
index 96f1608be..f4520a9cc 100644
--- a/contrib/ros/envs/example_launchers/trpo_gazebo_sawyer_push.py
+++ b/contrib/ros/envs/example_launchers/trpo_gazebo_sawyer_push.py
@@ -1,3 +1,5 @@
+import rospy
+
 from rllab.baselines.linear_feature_baseline import LinearFeatureBaseline
 from rllab.envs.normalized_env import normalize
 from rllab.misc.instrument import run_experiment_lite
@@ -10,6 +12,8 @@
 
 
 def run_task(*_):
+    rospy.init_node('trpo_sim_sawyer_push_exp', anonymous=True)
+
     env = TfEnv(normalize(PushEnv()))
 
     policy = GaussianMLPPolicy(
diff --git a/contrib/ros/envs/sawyer_env.py b/contrib/ros/envs/sawyer_env.py
index 51e79fec2..f0586bc8f 100644
--- a/contrib/ros/envs/sawyer_env.py
+++ b/contrib/ros/envs/sawyer_env.py
@@ -79,7 +79,6 @@ def __init__(self,
         Serializable.quick_init(self, locals())
 
         RosEnv.__init__(self)
-        rospy.init_node('SawyerEnv', anonymous=True)
 
         # Verify robot is enabled
         self._robot = Sawyer(