Modeling and evaluation of harbor crane work

Currently there is a strong demand of increasing the efficiency of the work chain in container harbors. Several methods aiming at improving the logistics of the container yard have been proposed. Nevertheless, the efficiency of an individual container crane has not been investigated, even though it has a great impact on the efficiency of the whole work chain. This paper proposes a method for modeling and evaluation of the work of a single harbor crane. In the method, the work is first divided into comprehensible tasks. A sequence of tasks produces a work cycle. The work cycle is modeled by a Hidden Markov Model (HMM). Each task of the work cycle corresponds to a state of the HMM. The HMM recognizes the states of the work cycle by using measurements readily available in the database of the crane. The resulting task time distribution directly reveals the bottlenecks of the work.