Christian Tummel

CloudLogistic – Line-Based Optimization for the Disposition of LTL Shipments

As a real application of the in Tummel et al. (The Multi-Depot Heterogeneous Fleet Vehicle Routing Problem With Time Windows And Assignment Restrictions (m-VRPTWAR), 2011) introduced problem class m-VRPTWAR – the “multi-depot heterogeneous fleet vehicle routing problem with time windows and assignment restrictions” – this paper will introduce the so-called “CloudLogistic” concept. The problem addresses the assignment of a set of shipments to a set of freight routes in order to minimize unused cargo volume of the vehicles. The assignment of each shipment is restricted to a subset of freight routes. Furthermore, the shipment has to be delivered in a specific time window. Therefore, it is necessary to determine an order of shipments for each freight route that guarantees the observance of all time windows. The problem class m-VRPTWAR abstracts the implied optimization problem. Besides the introduction of the “CloudLogistic” concept, the main requirements for the software-based shipment processing are discussed, which is the central part of a software-based solution for an implied freight cooperation of Less Than Truckload (LTL) shipments. For the evaluation of problem-specific solvers, as well as for an improved evaluation of the feasibility of the m-VRPTWAR, realistic test data come into place according to Tummel et al. (An Incremental Online Heuristic for Evaluating the Feasibility of the m-VRPTWAR, 2011). Besides a detailed description of the concept a method for the generation of realistic test data will be presented. Finally the evaluation of a Repacking First Fit approach (RFF) as a solution for the discussed feasibility check will be extended by considering different choices of repacking depths.

The Multi-Depot Heterogeneous Fleet Vehicle Routing Problem With Time Windows And Assignment Restrictions (M-VRPTWAR)

In this paper, the multi-depot heterogeneous fleet vehicle routing problem with time windows and assignment restrictions (m-VRPTWAR) is introduced. The problem addresses the assignment of a set of shipments to a set of freight routes so that unused cargo volume of the vehicles is minimized. The assignment of each shipment is restricted to a subset of the freight routes. Furthermore, the shipment has to be delivered in a specific time window. Thus, it is necessary to determine an order of the shipments of each freight route that guarantees the observance of all time windows. Firstly, a formulation of an integer linear program (ILP) for solving the m-VRPTWAR is developed and the problem is proven to be NP-hard. Afterwards, the ILP is evaluated by solving several large-scale scenarios using the solvers CPLEX and Gurobi.

An Incremental Online Heuristic For Evaluating The Feasibility Of The M-VRPTWAR

In this paper, a heuristic approach for evaluating the feasibility of an instance of the Multi-Depot Heterogeneous Fleet Vehicle Routing Problem with Time Windows and Assignment Restrictions (m-VRPTWAR) is presented. The heuristic approach tries to find a feasible solution by solving two sub-problems of the m-VRPTWAR: a 3-Dimensional Variable-sized Bin-Packing problem (3DVSBPP) and a Traveling Salesman Problem with Time Windows (TSPTW). In order to address the first problem, an online algorithm is presented. This algorithm expects shipments to be inserted in an online fashion and assigns them to freight routes by first-fit. Repacking already assigned shipments is permitted. The TSPTW part of the problem is addressed by a 2-phase heuristic that is based on algorithms presented in Savelsbergh (1985) and Ascheuer (1996). If there is no feasible solution for a new shipment, this shipment will be rejected. The heuristic approach is evaluated by solving several large-scale scenarios and monitoring the rejection rate of the algorithm.

Towards Measuring User Experience, Activation and Task Performance in Immersive Virtual Learning Environments for Students

The technological progress in the field of Virtual Reality (VR) facilitates the availability and applicability in learning scenarios. Applying VR technologies in a virtual environment (VE) can further intensify the students’ learning experience to a more immersive and engaging involvement in the learning process. In order to use immersive virtual learning environments (VLE) in high school and higher education as a teaching and learning tool, first it has to be assessed which individual-related variables influence the VLE experience and in which way the user experience (UX) affects task performance in VLEs. The paper describes the concept of a specific VLE using Minecraft as a setup for a research study. Furthermore, the research design is outlined. The paper describes the conducted pretest of the study and presents preliminary results of the pretest.

NEXT LEVEL BLENDED LEARNING FOR AN EXCELLENT ENGINEERING EDUCATION

The digitalization of higher education in general and Blended Learning in particular have been focused for a long time. Next to applying new technology, focusing on interdisciplinary competences during studies has always been part of engineering studies. The lecture Communication and Organization Development (KOE) is a constituent part of the Bachelor’s studies of the faculty of mechanical engineering, addressing 1500 freshmen of the RWTH Aachen University. Since decades the lecture KOE is part of frequently used revision along with an agile progression of applied mixed methods and new demands using the previously mentioned teaching development. The number of first semester students is still increasing and freshmen entering the university are partially minors (17 years old) due to the shortened school career. To master these and further challenges successfully, the didactical method mix is based on a cube model developed by Baumgartner and Payr (1996), which combines different levels and perspectives of a student-centred teaching approach. Elements such as the flipped classroom concept (Bretzmann, J. (2013): Flipping 2.0. Practical strategies for flipping your class, Bretzmann Group, LLC) (Bergmann. J., Sams, A. (2012): Flip your classroom. Reach every student in every class every day. International Society for Technology in Education, ASCD) along with the discussion forum, which were introduced during winter term 2013/2014, have become a permanent feature of KOE. Moreover, blended learning elements were increased by attaching a two-stage business simulation called ROBOFLEX to bring the theoretical elements to an immersive, problem-based learning approach. This paper presents the cube model’s current status as part of the method mix of KOE. Additionally, based on last year’s evaluation results, the paper examines student’s reactions towards the tutorial class ROBOFLEX as well as its potential for optimization. Furthermore a general analysis of the evaluation results from winter term 2013/2014 compared with 2014/2015 is performed to identify scope for the further development of KOE.

CloudLogistic – Geschäftsmodellentwicklung für eine Frachtenkooperation kleiner und mittlerer Straßengüterverkehrsunternehmen im Teilladungssegment

Dieser Beitrag adressiert einen ersten Entwurf eines Geschaftsmodells fur eine Frachtenkooperation von kleinen und mittleren Speditionsunternehmen des Strasenguterverkehrs im Bereich von Teilladungstransporten. Zunachst erfolgt eine kurze Einfuhrung, in der die Relevanz von Frachtenkooperationsbildungen kleiner und mittlerer Speditionsunternehmen aufgezeigt wird. Im Gegensatz zum Stuckgut- und Komplettladungsbereich existiert fur den Transport von Teilladungen allerdings bis dato noch kein optimiertes Logistikkonzept fur klein- und mittelstandische Transportnetzwerke im Strasenguterverkehr. Das CloudLogistic-Szenario wird als ein geeigneter Ansatz vorgestellt, um dieser Problemstellung zu begegnen. Den Kern dieses Beitrags bildet der Geschaftsmodellentwurf fur diesen Frachtenkooperationstyp mittels der Business-Model-Canvas-Methodik von Osterwalder/Pigneur, da dieser Ansatz gerade fur die Eingangsphase im Rahmen von Geschaftsmodellkonzeptionen bzw. bei Geschaftsmodellinnovationen geeignet ist. Der Beitrag empfiehlt, den Entwurf des Geschaftsmodells um eine geeignete Business-Architektur, die die relevanten Geschaftsprozesse im Hinblick auf die sich anschliesende Konzipierung der IT-Architektur fur die optimierte Zentraldisposition beinhaltet, zu erweitern.

Line-based optimization of LTL-shipments using a multi-step genetic algorithm

Motivated by the so-called “Cloud Logistic”-concept as an innovative, line-based way for dealing with less than truck load (LTL) shipments in cooperation networks, this paper introduces a genetic algorithm as a heuristical approach for dealing with multi-objective optimization problems. Based on the implied optimization problem - the NP-hard multi-depot heterogeneous fleet vehicle routing problem with time windows and assignment restrictions (m-VRPTWAR) - four different optimization goals of the “CloudLogistic”-concept are introduced and a multi-step approach is motivated. Therefore, two different optimization steps are presented and transferred into a genetic algorithm. Additionally, two innovative problem-specific genetic operators are introduced by combining a generation-based approach and a usage-based approach in order to create a useful mutation process. A further usage-based approach is used to realize a problem-specific crossover operator. The presented genetic multi-step approach is a useful concept for dealing with multi-objective optimization problems without the need of a single combined fitness function.

Potenziale für Smarte Innovation finden

Dynamisch und komplex – so ist unsere heutige Welt, weshalb es immer schwieriger wird, Innovationspotenziale zu erkennen und diese Innovationen dann systematisch zu gestalten. Haufig haben sich Strukturen und Prozesse schon so verfestigt, dass es schwer ist, das dort liegende Innovationspotential zu entdecken und zu nutzen (Strina/Jaime 2003). Nicht nur aus Forschung und Entwicklung (FuE) gehen Innovationen hervor, sondern jeder Prozess im Produktlebenszyklus kann und muss seinen Beitrag zum Innovationsprozess leisten. Die Entwicklung von Innovation ist eine permanente Herausforderung, die es entsprechend stetig neuen Anforderungen zu gestalten gilt (Isenhardt/Petry 2009).

Pump It up ! – Conception of a Serious Game Applying in Computer Science

Student attrition in mechanical engineering at German universities currently lies at about 40 %. A lacking sense of practical relevance for a future career are often named as reasons to quit studies. Over the past decade online games have become very popular for educational purposes. The approach of game-based learning, however, has proven to be suitable to motivate students. At RWTH Aachen University engineering students are imparted the relevance of computer science for their field through an e-learning environment including the online game Pump it up! The paper describes the conception and game design of the game including didactical and technical requirements related to it.

CloudLogistic – A Formal Description of Exclusive Freight Routes Based on a Multi-Zone Model with Elliptic Isochrones

Based on previous works this paper introduces a formal description of exclusive defined freight routes using a multi-zone approach to define source and target areas as elliptic isochrones. This work contributes to publications in context of the so-called “CloudLogistic” concept. As an innovative approach to strengthen the market position of Small and Medium-sized Enterprises (SME) and to increase their efficiency, the CloudLogistic concept aims for building a novel kind of freight cooperation in the area of Less than Truck Load (LTL) transports. The use of a multi-zone model enables the freight cooperation to maximize its efficiency, to implement innovative pricing strategies and to create individual unique selling features for each single Logistic Service Provider (LSP). By using an elliptic definition with geographical focal points for these zones a more flexible definition of source and target areas is realized. To reduce geographical impacts a time-based distance-function is used.