CRANE, A VEHICLE, AND A METHOD OF THE CRANE

A crane arranged to be mounted to a vehicle and including a crane controller. The crane controller includes a machine learning algorithm, including a neural network trained to calculate a movements of a crane components based on a received trajectory instruction. A supervisory controller is provided arranged to evaluate the calculated movements by comparing an accuracy measure to a predefined safety margin. The accuracy measure is defined by the difference between a crane tip position resulting from the calculated movements when moved from a current position, and a position of the crane tip calculated by the supervisory controller, based on the received trajectory instruction and on positions of individual joints of said crane boom system, when moved from the current position, and to generate an overruling operating instruction for a system of actuators, in response to the accuracy measure being larger than the predefined safety margin.