Minor wording change in paper

paper/paper.md

@@ -41,7 +41,7 @@ Collision-checking is an increasingly important tool as robots are deployed into
 
 Our package is similar to [Python-fcl](https://github.com/BerkeleyAutomation/python-fcl), which provides a Python binding of FCL that could also be used in a ROS architecture. The key difference is that our implementation is written in C++. The [ros_collision_checking](https://github.com/CoFra-CaLa/ros_collision_detection) package also offers a collision-checking system for 2D vehicles in a ROS environment. Our collision-checking system instead extends the general capabilities of FCL for proximity querying any geometric model and can thus be applied in numerous robotics contexts where such proximity information about the 3D environment is beneficial.
 
-The interface offered by `robot_collision_checking` is especially practical for obstacle avoidance and path planning in robotics use-cases. First, the robot's collision geometry (e.g., extracted from the geometrical ROS message types available to its URDF model) and any surrounding objects perceived by the robot's sensors (e.g., an Octomap representation of the environment given depth cloud data from an onboard RGB-D camera) can be added to a collision world, constructed and maintained using our `robot_collision_checking` interface. The resulting collision and distance information exposed by FCL through our package then enables safe plans to be generated for the robot's motion, e.g., using motion planning algorithms available to MoveIt [@coleman2014reducing]. The [constrained_manipulability](https://github.com/philip-long/constrained_manipulability) package demonstrates how to integrate robot manipulators with the `robot_collision_checking` package for such motion planning purposes.
+The interface supplied by `robot_collision_checking` is especially practical for obstacle avoidance and path planning in robotics use-cases. First, the robot's collision geometry (e.g., extracted from the geometrical ROS message types available to its URDF model) and any surrounding objects perceived by the robot's sensors (e.g., an Octomap representation of the environment given depth cloud data from an onboard RGB-D camera) can be added to a collision world, constructed and maintained using our `robot_collision_checking` interface. The resulting collision and distance information exposed by FCL through our package then enables safe plans to be generated for the robot's motion using standard motion planning algorithms, such as those available through MoveIt [@coleman2014reducing]. The [constrained_manipulability](https://github.com/philip-long/constrained_manipulability) package presents a few concrete examples of this procedure, demonstrating how to integrate the `robot_collision_checking` package into a motion planning framework for robot manipulators.
 
 # Acknowledgements
 Mark Zolotas is currently at Toyota Research Institute (TRI), Cambridge, MA, USA. This paper describes work performed at Northeastern University and is not associated with TRI.