Advanced Optimization Methods for Sustainable and Responsive Logistics
BayFOR 17.801
The project develops state-of-the-art optimization algorithms for energy-efficient and responsive logistics of express shipments, such as the collection of medical specimens for laboratory analysis, emergency supplies, on-demand deliveries of perishable goods or express deliveries in e-commerce. These methods combine advanced operational research methods and artificial intelligence techniques to work out concepts for sustainable and responsive logistics. The emphasis of the collaboration is to generate methodological knowledge at the cutting edge of digital technology with a high potential for significant spillovers.
Advanced Optimization Methods for the Management of Mobile Robots in Factories and Warehouses
BayFOR 16.328
In the modern industrial environment (digital factories and warehouses), automated mobile robots play an extremely important role in the performance of a large number of tasks, especially in the transport of parts and components between various work or service stations. Managing a fleet of automated mobile robots presents many challenges both in terms of safety and efficiency of their movements. The problem of planning the movements of automated mobile robots, whether in the environment of a factory or a warehouse, is therefore extremely complex. It can also be formalized as a difficult combinatorial optimization problem. It is therefore necessary to use state-of-the-art optimization methods to plan and manage in real time the movements of a fleet of automated mobile robots. The goal of this project is the development of such optimization methods.
INFORM-DECIDE – Multi-Stage Decision Making under Uncertainty in Data-Rich and Competitive Environments
Proposal for a Research Unit, DFG
[Translate to English:]
Incomplete information and uncertainty are inherent in transportation and logistics, with information constantly evolving due to factors such as road congestion, new transportation requests, availability of containers/trucks/vessels, and infrastructure decisions of competing mobility providers. In addition, related planning problems often require decisions spanning multiple time periods or even time scales, ranging from long-term infrastructure decisions coming with a time horizon spanning over years to short-term decision problems which are carried out sequentially over a much shorter time period (weeks, days up to minutes). This means that many planning problems in transportation and logistics are best represented as multi-stage optimization problems under uncertainty and competition. This project envisions the development of a holistic decision-support framework that unites the strengths of alternative paradigms – such as stochastic optimization, robust optimization, online optimization, and game theory – and the potential of data-driven methodological approaches. The overall objective is to generate both a deeper methodological understanding while also providing practitioners with new solution algorithms to address multi-stage problems more efficiently and according to need.
AI-Driven Multimodal Mobility Optimization for Rural Regions
BMDV 45KI01A011
Drawing on AI methods, the project investigates the potential of intermodal transport, with particular emphasis on drones and delivery robots, to enhance the quality and efficiency of medical care in rural areas.
Sustainable Personnel Planning in Highly Customized Assembly Lines with Work Sharing (SuPerPlan)
DFG 414225725
Work sharing refers to the organization of work, in which the responsibility for completing some tasks is shared between several actors. Their efforts must be coordinated and synchronized. Work sharing can significantly improve operations by reducing variance in station loads, e.g., as actors (workers or machines) at underutilized stations support bottlenecks. Yet, planning such systems is challenging. This project developed quantitative planning approaches to manage systems with work sharing.