diff --git a/Universal_Robots_ROS2_Driver-not-released.humble.repos b/Universal_Robots_ROS2_Driver-not-released.humble.repos
index f1f74842e..2883541e6 100644
--- a/Universal_Robots_ROS2_Driver-not-released.humble.repos
+++ b/Universal_Robots_ROS2_Driver-not-released.humble.repos
@@ -4,3 +4,7 @@
 # Once Upstream packages are released and synced to the target distributions in the required
 # version, the entry in this file shall be removed again.
 repositories:
+  ros2_controllers:
+    type: git
+    url: https://github.com/ros-controls/ros2_controllers
+    version: 2.14.0
diff --git a/Universal_Robots_ROS2_Driver.humble.repos b/Universal_Robots_ROS2_Driver.humble.repos
index 1ccf74a37..f9e948e30 100644
--- a/Universal_Robots_ROS2_Driver.humble.repos
+++ b/Universal_Robots_ROS2_Driver.humble.repos
@@ -14,7 +14,7 @@ repositories:
   ros2_controllers:
     type: git
     url: https://github.com/ros-controls/ros2_controllers
-    version: 2.12.0
+    version: master
   kinematics_interface:
     type: git
     url: https://github.com/ros-controls/kinematics_interface.git
diff --git a/ur_controllers/src/scaled_joint_trajectory_controller.cpp b/ur_controllers/src/scaled_joint_trajectory_controller.cpp
index acfe0bdb6..ae7616126 100644
--- a/ur_controllers/src/scaled_joint_trajectory_controller.cpp
+++ b/ur_controllers/src/scaled_joint_trajectory_controller.cpp
@@ -106,7 +106,7 @@ controller_interface::return_type ScaledJointTrajectoryController::update(const
   }
 
   JointTrajectoryPoint state_current, state_desired, state_error;
-  const auto joint_num = joint_names_.size();
+  const auto joint_num = params_.joints.size();
   resize_joint_trajectory_point(state_current, joint_num);
 
   // current state update
@@ -206,7 +206,7 @@ controller_interface::return_type ScaledJointTrajectoryController::update(const
         // send feedback
         auto feedback = std::make_shared<FollowJTrajAction::Feedback>();
         feedback->header.stamp = time;
-        feedback->joint_names = joint_names_;
+        feedback->joint_names = params_.joints;
 
         feedback->actual = state_current;
         feedback->desired = state_desired;