Work Note (Part I)

(versione italiana)

The work we are doing as consultants-researchers in the logistics field, emphasizes the existence of a gap among the industrial, the research and the academic world with respect to the management of logistics processes.
It exists a strong orientation to the operative time horizon, that usually forces the enterprises to adopt well-established solutions rather than consider innovative ones. In fact, these solutions could require a greater analysis and modelling effort. On the other hand, in the research field, the lack of data addresses studies based on realistic data sets and produces sophisticated models, often too distant from the reality.

Why using analysis and model formulation?
Analysis allows the understanding of the AS-IS of the logistics processes, the detection of criticalities and the definition of the main aspects for meeting requirement of the operative phase. After the analysis, the modelling phase allows the representation of the problem in a mathematical way, in order to solve it by using suitable tools. Typically, the researchers focus on the modelling phase while studying methods, models, decision-making tools for several fields as logistics one.
It is interesting to notice that the analysis and modelling efforts that an enterprise should carry out for bring in the innovation and the optimization of its logistics processes, produces a strong positive impact on the effectiveness of the processes that often corresponds to a costs decreasing.
Mainly, the logistics problems are represented by using combinatorial optimization models, for instance:

• vehicle routing problem – for distribution and collection processes
• scheduling problem – for production processes
• auction problem – for procurement processes

Critical Issue
One of the main issue in the analysis phase is the lack of a common language among the enterprise work-force and the researcher and processes analyst. This fact often corresponds to a difficult in defining the process constraints and goals.
From the modelling point of view one of the main issue is the huge data and operative constraints to be modelled that often arouse a huge computational effort.