Past projects

The project focuses on an intra-organizational sales and operations planning problem. Operations persons are responsible for different production sites and have private information on production costs and capacity. Sales persons are responsible for different customer markets and have private information on customer demand. Due to function-oriented bonus schemes, both parties have conflicting interests: While operations persons aim to minimize production and inventory costs, sales persons aim to maximize turnover. That is why both parties have an incentive to misrepresent their private information in favor of their own interest. This behavior leads to inefficient capacity allocation and thus reduces the profitability of the company. To address this problem, we develop coordination mechanisms, which consider asymmetric information and the participants’ reluctance to share their private information. We aim to design efficient and incentive compatible capacity allocation mechanisms. Frank Hage is working on this project.

“Reuse of water in the food and bioprocessing industry” (REWARD) project, is a large interdisciplinary research effort with partners from Denmark, Netherlands and Germany aimed at developing methodologies and techniques for pro-active industrial water management. The project will combine state-of-the research in real-time monitoring, Nanosensor development, optimal process design, water microbiology and production & supply chain management to bring food and bioprocessing industries closer to self-sustainability - the closed factory principle - where water intake is diminished by re-using production streams. REWARD focuses on process water cases from the food and bioprocessing industries with multi-product lines and includes technology providers working with generic principles adaptable to other industries with related challenges such as the brewing industry, vegetable/fruit processing, bio-refineries and the fermentation industries. TUM School of Management along with Wageningen University together constitute the working package on water supply chain management that specifically focuses to find ways to integrate water reuse in methodologies for operations and supply chain management in food production. This is an essential part of any initiative to minimize water consumption or maximize water reuse opportunities. Mr. Sai Jishna Pulluru from TUM School of Management is responsible for the Scientific content of the water supply chain management working package. Sai Jishna Pulluru is working on this project.

The increasing vehicle heterogeneity is pushing the widespread mixed-model assembly line to its limit. In alternative flexible layouts, the assembly stations are no longer arranged in a serial manner and no longer linked by a paced carrying system but rather by automated guided vehicles. Such flexible layouts can improve makespan, station utilization and disruption risk resilience when a large variety of vehicles is produced in low quantities. However, the advantages come at the expense of increases in space requirements as well as production and transportation complexity. Our objective is therefore to analyze the circumstances under which flexible layouts are in fact superior to mixed-model assembly lines. In this research project, we study the strategic layout design as well as the operational scheduling of flexible layouts. We compare the key performance indicators of flexible layouts to mixed-model assembly lines for various production scenarios to determine which type of layout is most suitable for which type of scenario. As research methodologies, we apply mixed-integer linear programing as well as advanced solution algorithms and heuristics.

The mid-term production planning in semiconductor manufacturing is confronted with a large number of challenges: the manufacturing network is in most cases flexible, alternative, non-identical machines have to be considered, the cycle time is a non-linear function of utilization and the lead-times of the products are long. The current research addresses some of the issues by trying to accurately represent capacities, reduce nervousness between plans in consecutive planning periods and by appropriately considering prioritized demand classes and load-dependent cycle times.

Dried dairy products are a popular ingredient in the production of various food products such as chocolate, cheese and yogurt. Milk and whey are concentrated in large volumes and further dried to powders using enormous amounts of energy. This project investigates if it is possible to omit the energy intensive drying process and instead produce only concentrates with more efficient concentration processes. Avoiding the drying process can save as much as 1000 GWh/yr in Germany alone. Therefore the Federal Ministry of Food, Agriculture and Consumer Protection has granted a funding to the project.

The objective is to develop innovative methods for stabilizing concentrates and new forms of logistics, such that the concentrates can be transported and stored like powders, i.e. for sufficient time and under similar temperatures. In an interdisciplinary approach we will investigate the required changes in logistics and processing, and evaluate the concept as a whole with regard to its cost-effectiveness and sustainability. Finally, all processing steps along the supply chain are validated through an assessment stage at industrial scale together with our project partners. The project is carried out in cooperation with the chair for Food Process Engineering and Dairy Technology as well as two project partners from industry, Privatmolkerei Naarmann GmbH and Bayerische Milchindustrie eG.

The proliferation of products has become a common phenomenon in today's business world. Options are widely used by many car manufacturers to provide customization possibilities and capture the surplus of customers with heterogeneous quality valuations, tastes or budget constraints. However, their introduction can lead to a large numbers of theoretically feasible product variants. The offering of options as part of option bundles could limit the increase of product variety. 

The project focuses on the development of an analytical method for the allocation of a set of options to a required number of bundles and on the analysis of the effects of option bundling on product variety. To this end, a set of bundle candidates is derived using discrete customer choice models and past purchases information. The required number of bundles is selected from the candidates set by a revenues maximizing mixed-integer linear program. In order to ensure the practical relevance of the framework, various bundling policies and bundle discounts are considered.

The objective of project is to develop a mixed-integer linear optimization model to determine market specific volumes, as well as the variant mix to be used for the market introduction and the ramp-up of a new car model. Various ramp-up situation characteristics in production and supply chain, like maturity levels, complexity, model-mix restrictions or variants sequences, are considered. In order to reflect the allocation situation, fairness considerations are added to the optimization model.

The focus of the project lies in the coordination of the operations for a multi-plant production network and a 2 years planning horizon in order to minimize total production and logistics costs. Production, inventory transportation and setup costs are considered in the framework. The coordination must consider fixed external demand, time constraints and high setup costs. In order to derive such a plan, a methodology based on a network flow formulation and a disjunctive graph model is considered. A case study with numerous experiments is being developed.

Robot cells, also known as cluster tools, are used in semiconductor manufacturing to produce silicon wafers. For such equipment, the minimization of cycle time is a paramount issue. The project focuses on the derivation of a schedule to minimize the cycle time in a robot cell, as well as on the development of quantitative approaches for obtaining optimal schedules. Complex scheduling requirements, such as stochastic processing times, wafer delay constraints, reentrant wafer flows and/or diverse architectures have to be considered. A Petri Net modelling of the robotic cells is used in order to conceive schedules.

Master planning in semiconductor manufacturing assigns hard orders and forcasted demand on finished product granularity into supply chain resources. Simultaneously, the Available-to-Promise (ATP) plan is generated for forecasted demand and forwarded to order management on aggregated product granularities. Under certain conditions, when single products face capacity restrictions, real time order promising results in unreliable, too early delivery dates. The project focuses on the development of different approaches for an accurate order promising process with aggregated product ATP. A simulation study will be run with the developed approaches on the example of a semiconductor company.

The focus of the project is to develop a production and inventory plan prior to market launch in order to cut lost peak sales while managing uncertainty and to improve short-term capacity planning by considering ramp-up and different learning effects. To this end, the key challenges for new product introduction in the pharmaceutical industry are identified. Production and inventory build-up plans are derived in all echelons prior to market launch under uncertain market conditions. Additionally, appropriate short-term capacity and ramp-up of the initial production are identified while considering several learning dynamics.