Optimizing procurement in a Supply Chain for improved logistics performance

[laro]

Optimizing procurement in a Supply Chain for improved logistics performance
Prof. Dr.-Ing. habil. W. Dangelmaier, Dr. Christoph Laroque

Today, planning of production and transportation in a Supply Chain is done separately, i.e. all suppliers make their production plans without considering how the products will be shipped. It is commonly assumed that transports are available at low prices anytime. That results in production plans that take no logistical requirements in account and thus leads to suboptimal tours and trucks that are not fully loaded, and ultimately to more traffic, higher costs, and additional environmentally harmful emissions. It is obvious that an integrated planning approach looking at the Supply Chain rather than at single companies can provide better plans.
Integrated planning has been proposed before, but only on the distribution side. The advantage is that in this case all operations are under control of a single company. A possible scenario consists of a sweater company with different customers. If only production parameters are considered, the resulting production sequence will go from light colors to darker ones in order to save on cleaning processes. However, this sequence might not be optimal for tour planning so that the resulting trips are relatively long. If the sequence is altered in a way such that orders for customers in one region are produced in succession, better tours can be planned.
In contrast to that, integration on the procurement side means the integration of several different companies which are all legally independent. Accordingly, the problem itself is also more complex because it encompasses numerous production processes. Therefore, transportation improvement can only be realized if all production processes are aligned. Under the assumption that customer deadlines have to be respected, it might be reasonable to bring forward some orders and to postpone others that originally had a time window for delivery.
The downside to this approach is that it will most likely lead to higher stock levels, which represent a break with the current paradigms of lean manufacturing and just-in-time delivery. Yet in a time of rising oil prices and consequently high costs for transportation, a reassessment of the target system in production and logistics has become necessary.
The aim of this work is twofold. First, a quantifiable estimation of the benefits of integrated planning has to be made. This represents a best case scenario and all stakeholders are planned centrally. In the next step, methods allowing practical use have to be devised. Due to the fact the companies consider their internal data such as capacities to be confidential, collaboration has to be organized by proposals, counter-proposals, and incentive offering. An implementation using a multi-agent system seems most promising as this is a modeling technique which fits the requirements of decentralized planning found in a Supply Chain very well. In the end, the expected results of this work are both centralized and decentralized planning models and the respective savings in terms of money and emissions.
[b]