The use of overhead cranes in warehouses and
factories has advantages for handling and transporting bulky
and/or heavy loads. But it also involves risks such as collisions with
other fixed or mobile elements in the working environment.
Different types of sensors have been used for monitoring its
operation, mainly artificial vision. In this paper, it is employed a
three-dimensional (3D) LiDAR to capture the workspace of a bridge
crane. The point clouds generated by this laser sensor are delivered
to a convolutional neural network to detect the position of the bridge
and its carriage, which allows to locate the hook and the suspended
load afterwards. Additionally, the laser scans can also be used to
warn the operator of possible collisions with fixed elements of the
warehouse. The tests carried out show that the proposed system can
be successfully used for monitoring overhead cranes.
