Research results from the ongoing PhD projects are regularly presented at the PhD Workshop Northeast.
Transport optimization in empties container networks of Volkswagen AG
|Duration:||Ongoing (as of Dec. 2020)|
Volkswagen Konzernlogistik GmbH & Co. OHG is responsible for planning, controlling and monitoring the transport networks in the Volkswagen AG group network. This includes the transport of externally manufactured components from suppliers to the Group-wide production plants, which is carried out almost exclusively in returnable load carriers due to the fact that transporting individual, loose components makes no sense from either an economic or a logistical point of view. After arriving at the plant, the loaded load carriers are lined up as needed on the production line or transported just-in-time to the production line before the parts are removed for assembling the vehicle. In order to close the load carrier cycle and to ensure that future component deliveries can also be made in the returnable load carriers, it is imperative that the empties are transported back from the plants to the suppliers. These transports, which do not directly add value, are cost-intensive and cause high CO2 emissions. Last but not least, as part of the Group strategy "TOGETHER 2025+"and the derived goTOzero strategy, Volkswagen has committed itself to reducing CO2 emissions by 30% by 2025 compared with 2018 and to achieving CO2 neutrality in its balance sheet by 2050.
The aim of this dissertation project is to plan the transport of empties from plants to suppliers as efficiently as possible with regard to current and future requirements. In close cooperation with Volkswagen Group Logistics, new concepts for load carrier allocation in the transport network are to be developed and compared using an experimental performance analysis based on realistic test instances. The main challenge is the sheer size of the optimization problem on the one hand and the complexity of the process structures on the other hand. For example, the transport problem to be solved involves not only around 50 plants, more than 1,000 suppliers and around 50 universal load carrier types, but also shipment lead times, degressive transport costs and different modes of transport. Due to the fact that the underlying Dynamic Multicommodity Fixed Charge Transport Problem belongs to the class of NP-hard problems in combination with the size of the instances to be solved, standard solvers reach their limits. Therefore, the main focus of the research project shall be the development of heuristic solution approaches. In a first step, a realistic instance set for the problem of transporting empties will be generated. In the next step, existing network approaches are to be taken up for the problem structure at hand and new approaches are to be conceptualized before solution procedures are developed for the generation of best possible feasible solutions. The focus here is on the development of construction heuristics specific to the problem. Subsequently, the solutions generated this way are to be further improved by the use of suitable improvement procedures.
Route planning of internal roll-off container logistics
|Duration:||Ongoing (as of Dec. 2020)|
The global transport volume has been growing continuously for years. According to Statista, the transport volume in 2019 amounted to more than 4000 million tons in Germany alone. The majority of these transports were carried out by truck on the road. In addition to the public space transport, transports also take place in-plant. The transport of goods does not create any value itself. Its necessity, and thus its value, result from the availability of a good at a certain place at a certain time. Efficient management of existing transport needs therefore holds potential for saving costs and improving the performance of a transport system.
Roll-off container transport systems can be used in intermodal transport. Roll-off containers are load carriers with rollers and a hook to hold them. They enable fast handling without stationary loading aids. Their use in the mining industry is suitable due to the intermodal transport of goods above and below ground by transport vehicles and their shaft transport. While the lower frame of the container is standardized, different superstructures open up a wide range of applications. Examples include superstructures for bulk goods, flexible loading areas with hinged doors, tanks and machine superstructures.
Together with K+S AG, we are looking at planning tasks in the area of roll-off container transport at the Zielitz site. All transport activities are covered, from the delivery of empty roll-off containers to the transport of loaded roll-off containers to the servicing of material deliveries and collections with roll-off containers. The goal is to develop holistic planning for the efficient execution of accruing transport requirements.
The practical application of computer-aided solution methods requires short computation times with good solution quality. Therefore, heuristic solution algorithms are developed within the project.
Medium-term planning of flexibility tools in flow line and assembly line production using the automotive industry as an example.
|Start:||August 1, 2005|
In saturated triad markets, the market demands on the automotive industry are increasing: The customer expects attractive products in all vehicle classes, so that even in the premium segment, additional prices can only be achieved through customer-experienced added value. Manufacturers are reducing product life cycles, producing niche products, and increasing the degree of innovation and the number of variants that customers can experience. On the development side, production-oriented product development and the associated modularization and standardization concepts ensure that complex products can be produced at acceptable costs. In production, cost pressure generates a consistent value-added orientation.
However, a prerequisite for high returns on sales for manufacturers is high utilization of production lines, which in increasingly dynamic and uncertain markets can only be achieved through a high degree of flexibility. On lines that are flexible in terms of design for the volume market, niche products can thus be integrated without difficulty alongside highly standardized products. In the premium segment, this is more difficult, as efficient production of vehicles with a large number of variants and widely differing specifications is only partially possible on lines that are only flexible in terms of series. For this reason, other flexibility concepts are gaining importance for these manufacturers in addition to the series-flexible line.
Within the framework of a research cooperation with the Research Center Ulm of the DaimlerChrysler AG, the possibilities of planning flexibility instruments within a planning period of approx. 3-5 years are investigated.
Taking into account limited buffer capacities between the individual trades and diverse technical restrictions, the main focus of the research is on medium-term personnel requirement planning. Taking into account legal, collective bargaining and internal company regulations, shift and break planning in the face of fluctuating labor demand is combined with the concept of learning curves. The fluctuations in manpower requirements are mainly due to the non-constant demand for end products and the resulting need to adjust working times and production speeds on the individual production lines. The learning curve concept is suitable for mapping accelerations through regular practice of newly learned work processes after a reconfiguration of a production line.
In a further step, the interdependence of production lines, which already exists in the context of personnel (deployment) planning, is supplemented and reinforced by the possibility of distributing the production program to parallel production lines.
The developed procedures will be tested on suitable case studies and will then be integrated into the internal software environment of the cooperation partner for production planning and control.