Motion Planning of a Cable Array Robotic Crane on Board a Naval Vessel
Problem Description
In an attempt to optimize time efficiency, the U.S. Navy loads cargo
on to cargo vessels in random order. The cargo container which is
closest to the loading point is placed on deck in no particular
location. After the ship is fully loaded, it is desired to sort the
cargo in a specific order while on route to the prescribed
destination. Usually in a war time situation it is desired to get the
weapons unloaded before supplies such as food. Therefore the
containers must be sorted so that the ones with weapons will be
located on the side of the deck which will meet with the port dock. A
new crane called the cable array robot has been proposed to the Navy
for the purpose of loading, unloading and sorting cargo on deck. This
project will analyze the problem of sorting the cargo in to a
specified order and how the sorting process will effect the motion
planning of the robot.
Motion Planning Specifics
When the ship is loaded, the environment is known and each object is
identical. The containers are parallelepipeds and can only have one
orientation. The ship deck can be treated as a grid where there are
only discrete possible locations for the containers. The problem
begins with a list which specifies the desired locations and current
locations of each container. The problem is to move the containers in
a specific order so that performance criteria such as time or distance
are minimized. We will also consider other optimization parameters
such as to minimize the movement of the crane within areas of the
workspace which may be dynamically unstable due to the motion of the
ocean. If the natural frequencies of vibration of the crane are
matched by the sea, resonance will occur. Therefore areas of the
workspace which have natural frequencies similar to the ocean can be
avoided in the motion planning. The environment is time varying since
the containers are moving but it is totally predictable since the
robot is the controller of the environment. Therefore the environment
is certain but there is a complex optimization path planning problem.
The paper form our class list which is most suited for this project is
"On Motion Planning in Changing , Partially Predictable Environments"
LaValle and Sharma. The results of this project will include a
heuristic approach to the planning along with a more rigorous solution
to the optimization of path along the lines of LaValle and Sharma.
Several different optimization criteria will be analyzed.