Universität Bremen
Fachbereich Physik, Elektro- und Informationstechnik
Institut für Telekommunikation und Hochfrequenztechnik
Arbeitsgruppe Kommunikationsnetze
Otto-Hahn-Allee 1,
28359
Bremen,
Germany
Tel:
+49 421 218 62365,
Fax:
+49 421 218 9862365
E-Mail
, Homepage
B1 - Entwicklung von Methoden zum selbststeuernden Routing autonomer logistischer Objekte in Transportnetzen
Prof. Dr.
Carmelita
Görg
Prof. Dr.-Ing.
Bernd
Scholz-Reiter
Motivation
By research on autonomously controlled routing of autonomous logistic objects, based on methods from communication networks and other areas, subproject B1 represents the core idea of the CRC 637. The overall vision of the CRC consists of autonomous, decentralized control systems for logistic processes. This leads to a decentralized coordination of autonomous logistic objects in a heterarchical organization. The potentials and limits of a paradigm change from central planning and control systems towards an autonomous control of logistic processes are investigated in the subprojects in different areas of logistics. Subproject B1 here represents the basic idea of autonomous control: the autonomous decision making and routing of logistic objects in dynamic environments. By the research of subproject B1, methods will be available for autonomously controlled routing of logistic objects along a logistic process chain, consisting of production, assembly and transport processes, at the end of the 3rd phase.
Results Phase 1 (2004-2007)
In the first phase of the CRC 637, it was investigated if and how algorithms from communication networks can be applied to logistic problems. The area of transport logistics was first investigated here. The general transferability and applicability of routing principles from communication networks was proven and the Distributed Logistics Routing Protocol (DLRP) was developed as autonomous control method for the area of transport logistics. In an evaluation by comparison to conventional central optimization methods, the DLRP has shown competitive and in some cases better performance while the autonomous control additionally has advantages outside the optimization targets, e.g., with respect to flexibility and robustness. Therefore, at the end of the first phase, a method for autonomous control of vehicles and goods was available, whereby a basic implementation of the CRC 637 vision for transport logistics was achieved.
Objectives Phase 2 (2008-2011)
In this subproject's continuation, it will be investigated how the autonomously controlled routing of goods and means of transport is to be adapted under extended close-to-reality requirements. These requirements concern practically relevant restrictions and complex structures. Further, the potentials and limits of the basic concepts, developed in the first project phase, will be explored. These consist of three main points: consideration of load carriers, routing methods for the scalability to large networks, and informational barriers which exist in practice.
Another important aspect, especially for the recognition and evaluation of limitations, is the creation of realistic benchmark scenarios. The concepts for autonomous logistics are so different from the traditional transport logistic approaches that established benchmarks (such as Solomon's scenarios) cannot be applied here.
Another aspect is the transfer of the DLRP-concept, which is efficient for transport logistics, to other fields of logistics.
Results Phase 2 (2008-2011)
In the second phase, the autonomous control method DLRP was extended for close-to-reality situations and its suitability was shown. This included extensions and limitations for the scalability in large logistic networks, investigations on limitations and effects of technical and business strategical constraints in the information flow and extensions of the DLRP concept by load carriers. The functionalities for business strategical constraints extend the DLRP for simultaneous use of multiple, independent companies in contrast to isolated planning for individual fleets.
In addition to DLRP for transport logistics (DLRPt, t=transport), the DLRP concept was transferred to production logistics (DLRPp, p=production) and has proven advantages in comparison to central control methods here. For both areas, corresponding benchmark scenarios were developed and used. Further, a successful transfer back to communication networks has shown the great potential of DLRP for other areas than logistics. After the second phase, a DLRPt for transport and a DLRP for production is available for autonomous control of logistic processes.
By research on autonomously controlled routing of autonomous logistic objects, based on methods from communication networks and other areas, subproject B1 represents the core idea of the CRC 637. The overall vision of the CRC consists of autonomous, decentralized control systems for logistic processes. This leads to a decentralized coordination of autonomous logistic objects in a heterarchical organization. The potentials and limits of a paradigm change from central planning and control systems towards an autonomous control of logistic processes are investigated in the subprojects in different areas of logistics. Subproject B1 here represents the basic idea of autonomous control: the autonomous decision making and routing of logistic objects in dynamic environments. By the research of subproject B1, methods will be available for autonomously controlled routing of logistic objects along a logistic process chain, consisting of production, assembly and transport processes, at the end of the 3rd phase.
Results Phase 1 (2004-2007)
In the first phase of the CRC 637, it was investigated if and how algorithms from communication networks can be applied to logistic problems. The area of transport logistics was first investigated here. The general transferability and applicability of routing principles from communication networks was proven and the Distributed Logistics Routing Protocol (DLRP) was developed as autonomous control method for the area of transport logistics. In an evaluation by comparison to conventional central optimization methods, the DLRP has shown competitive and in some cases better performance while the autonomous control additionally has advantages outside the optimization targets, e.g., with respect to flexibility and robustness. Therefore, at the end of the first phase, a method for autonomous control of vehicles and goods was available, whereby a basic implementation of the CRC 637 vision for transport logistics was achieved.
Objectives Phase 2 (2008-2011)
In this subproject's continuation, it will be investigated how the autonomously controlled routing of goods and means of transport is to be adapted under extended close-to-reality requirements. These requirements concern practically relevant restrictions and complex structures. Further, the potentials and limits of the basic concepts, developed in the first project phase, will be explored. These consist of three main points: consideration of load carriers, routing methods for the scalability to large networks, and informational barriers which exist in practice.
Another important aspect, especially for the recognition and evaluation of limitations, is the creation of realistic benchmark scenarios. The concepts for autonomous logistics are so different from the traditional transport logistic approaches that established benchmarks (such as Solomon's scenarios) cannot be applied here.
Another aspect is the transfer of the DLRP-concept, which is efficient for transport logistics, to other fields of logistics.
Results Phase 2 (2008-2011)
In the second phase, the autonomous control method DLRP was extended for close-to-reality situations and its suitability was shown. This included extensions and limitations for the scalability in large logistic networks, investigations on limitations and effects of technical and business strategical constraints in the information flow and extensions of the DLRP concept by load carriers. The functionalities for business strategical constraints extend the DLRP for simultaneous use of multiple, independent companies in contrast to isolated planning for individual fleets.
In addition to DLRP for transport logistics (DLRPt, t=transport), the DLRP concept was transferred to production logistics (DLRPp, p=production) and has proven advantages in comparison to central control methods here. For both areas, corresponding benchmark scenarios were developed and used. Further, a successful transfer back to communication networks has shown the great potential of DLRP for other areas than logistics. After the second phase, a DLRPt for transport and a DLRP for production is available for autonomous control of logistic processes.
Project Staff
