SynchroTrans: Multi-Dimensional Synchronisation of Heterogeneous Resources in Transport

• Applicant: Professor Dr. Stefan Irnich
• Subject Area: Accounting and Finance
• Term: from 2010 to 2018
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 178444147

The project SynchroTrans aims at providing scientific progress for modeling and solving (heuristically or exactly) vehicle-routing problems characterised by multiple synchronization constraints (Routing Problems with Multi-dimensional Synchronization, RPMS).RPMS are characterised by the fact that for fulfilling a transport request several resources need to be synchronised in multiple ways. In this context, resources are for example vehicles, drivers, means of transport etc. Request-specific synchronization is the basic question of assigning requests to different vehicles as known in the standard vehicle-routing problem. Besides, RPMS require synchronization with respect to temporal, spacial, and load-specific aspects.The fundamental hypothesis to be confirmed by the SynchroTrans project is the following: By using tailor modeling approaches and integrated solution algorithms it is possible to significantly better solve vehicle-routing problems with multiple synchronization constraints either in an exact or a heuristic manner. This enables the use of quantitative planning tools and decision support systems in corresponding applications in transport and logistics.The basic questions to be answered within the SynchroTrans project are the following:1. In which areas and applications can we find planning tasks that can be modeled and solved as vehicle-routing problems with synchronization constraints?2. What is a generic and useful graph-based representation of RPMS to provide a basis for an exact MIP-based algorithm?3. What are powerful construction methods for computing start solutions to RPMS?4. What are metaheuristic procedures for RPMS that handle synchronization aspects in an integrated way?5. What are the right methods to takle the problem of interdependencies between resources in local search-based components?6. How can we integrate robustness aspects in formulations for RPMS?7. How can we solve the subproblem in branch-cut-price algorithms for RPMS efficiently?8. What are special cases of RPMS that can be modeled and solved with tailored solution procedures?The most important work packages are:1. a comprehensive literature review of RPMS applications2. the development of a generic, graph-based modeling concept for general RPMS3. the implementation of a branch-cut-price algorithm for RPMS4. the implementation of metaheuristics for solving RPMS

• DFG Programme: Research Grants