Objectives

The goal of the project is to define, implement and assess a dynamic integrated system to solve a wide range of complex distribution and logistics problems. These problems include routing of multiple delivery vehicles from depots to locations. The situation can be handled at the strategic level where baseline routes are created based on probabilistic demands over a long period of time. At the operational level, these routes are adapted to fit every-day demands. Finally once having left the warehouse, trucks may be re-routed in the case of break-downs or unexpected traffic problems.

For the end user, DYNALOG brings the following benefits

• Compute baseline routes that need not be changed very often, using probabilistic needs obtained via a demand forecasting module and changes are often costly for the company
• Able to stop taking orders closer to truck dispatch time, since routing are computed very quickly. This increases customer satisfaction and reduces the need of special produces to cope with late orders
• Able to modify routes in near real-time in order to react to events that occur during delivery. E.g. breakdowns, delays at customer sites.

The effective planning of logistics results in costs savings, increased competitive edge and greater customer satisfaction.

Expected Results

The expected results of the project include:

• A software based on innovative Constraint Programming Technology and Stochastic Computational Models to solve the problems of baseline routing, operational routing and dynamic re-routing
• To drive the baseline router, a demand forecasting module will be implemented
• A hardware implementation allowing users to make use of different available computers to solve large instances of these problems
• An integrated active decision system based Client-Server technology
• Assessment of these techniques in real situations

Scientific Approach

In order to be able to handle large networks (in excess of a few thousand nodes), we will rely on state-of-the-art Constraint Programming Technology and High Performance Computing and Networks.

Over and above usual routing models, DYNALOG will consider time distances that depend on the time of day of the trip and will model goods loading within the truck so as to be certain that goods can actually be loaded onto trucks.

Partners

The end-users in the project (MEVGAL, HOLYCON and FIEGE) have complementary problems that reflect the diversity of distribution needs across Europe.

• MEVGAL produces and delivers high-volume ultra-fresh dairy products such as fresh milk and cheeses and yoghurts. The special characteristic of this industry is that demand isn't always known when the trucks leave the warehouse.
• HOLYCON distributes a wide range of 'niche' dairy products to various types of customers. The ability of the DYNALOG system to fulfil orders whilst minimising the wastage due to out-of-date stock would lead to an increase in the cost-effectiveness of supplying ultra-fresh produce.
• FIEGE is a storage/distribution agency established throughout Eastern and Occidental Europe offering large volume logistic services : some 3000 loads (or 6 million tons of goods) per day handled. The DYNALOG system is expected to assist FIEGE by enhancing planning and by eliminating time constraints.

The supplier partners are

• COSYTEC will be the exploitation partner in the consortium. They have extensive experience in building and tailoring decision support systems to customer needs, and exploiting the results of EC projects. COSYTEC will also use its Second Generation Constraint Programming Technology to provide the deterministic-needs optimisation module.
• HPCN expertise will be supplied by NTUA and the PAC. In particular, NTUA will take responsibility for the stochastic optimisation module. The PAC will transfer parallel expertise to the exploitation partners with respect to the deterministic router and demand forecasting tool.
• Logistics expertise will be provided by OIS. With 10 years of experience in this area, OIS will be able to bridge the gap in needs between the end-users.

For more information

The Project Manager is Mr. Kostas Georgakidis from MEVGAL, and the coordinates are

(30) 391 59107
(30) 391 52901
kgeorgakidis@mevgal.gr

COSYTEC contact: info@cosytec.com