Autonome Roboter in der Intralogistik: Möglichkeiten zur Optimierung der Auftragsverteilung

Translated title of the contribution: Optimization of Task Assignments for Autonomous Transportation Robots

Erik Stefan Himmelsbach

Research output: ThesisMaster's Thesis

170 Downloads (Pure)

Abstract

OpenShuttle is a new product by KNAPP Systemintegration GmbH, based on a fleet of autonomous robots performing intralogistic transportation tasks. Due to their laser-based navigation they do not need additional infrastructure for moving through the warehouse and are therefore a highly flexible and easy-to-install alternative to conventional transportation systems. One of the company’s current projects involves the use of OpenShuttle at the distribution centre of a food retailer where the transport of pallets between different stations is necessary. However, the current OpenShuttle-system’s throughput is not sufficient to meet this project’s requirements. The goal of this master thesis was to increase the throughput by optimizing the task distribution to the robots. For this purpose it was necessary to first expand the current simulation by the battery status of the robots to approximate their charging behaviour in the simulation. Moreover, the previous task distribution algorithm has been expanded into a new, two-staged distribution algorithm: first a region is assigned to a robot, and only when the robot reaches its assigned region, then it is given the exact task. Based on this preparatory work the task assignment has been optimized to increase the throughput of the entire transportation system. This new task assignment also takes into account that every order has to be executed within a certain time frame. Furthermore, the assignment considers the current state of battery charge of the robots. This new task assignment algorithm estimates in advance the time it takes a robot to execute a certain order. This estimation is based on empirical values which were collected by measuring the time needed to execute previous orders. The algorithm chooses the assignment for which the time of travel to the first station and the waiting time are as low as possible. A comparison of the previous and the new task assignment showed that the mean execution time of an order with the new task assignment is by 13,6% lower than with the previous task assignment.
Translated title of the contributionOptimization of Task Assignments for Autonomous Transportation Robots
Original languageGerman
QualificationDipl.-Ing.
Supervisors/Advisors
  • Auer, Peter, Supervisor (internal)
Award date24 Oct 2014
Publication statusPublished - 2014

Bibliographical note

embargoed until 16-09-2017

Keywords

  • autonomous robot
  • task assignment algorithm
  • intralogistics

Cite this