@misc{HerrmannDilefeld, author = {Herrmann, Frank and Dilefeld, Maximilian}, title = {Improving the scheduling of AGV and AMR using simulation}, series = {OR23, International Conference on Operations Research, August 29 to September 1, 2023, Hamburg}, journal = {OR23, International Conference on Operations Research, August 29 to September 1, 2023, Hamburg}, abstract = {Automated Guided Vehicles (AGV) and intelligent Autonomous Mobile Robots (AMR) are an important tool for designing a flexible Industry 4.0 shopfloor. Both types of Mobile Robots are usually deployed with a Fleet Management software to manage (transport) orders and vehicles together with other superordinate control systems. The planning tasks in the superordinate control system used by both AGVs and AMRs can be divided into Material Flow Management, Job Management, Vehicle Dispatching and Job Processing. Requests for supply and removal of material is issued by the stations and has to be combined to transport jobs. Alternatively, more complex system often user external Material Flow Controllers (e.g. MES). Jobs are collected in a queue and their execution order is determined. Vehicles which can execute the next job have to be identified and the best one is selected. Jobs are split into smaller tasks which are sent to the Mobile Robot and executed. For AGVs navigation functions are mostly integrated into the Fleet Manager as part of the Job Processing, while AMRs perform lower-level navigation functions directly on the vehicle and use the superordinate control system for implementing restrictions and traffic control functions. A need for further research and development was identified especially for decision-making algorithms within the Fleet Management. The goal is to compare existing and develop new strategies to solve the different planning problems. For the individual planning tasks different techniques have been discussed and used in research and practice. However, there is no universal best solution and different options must be compared in the context of the specific use case.}, language = {en} } @inproceedings{DilefeldClausTeich, author = {Dilefeld, Maximilian and Claus, Thorsten and Teich, Enrico}, title = {Robot Task Assignment In Dynamic Factory Environments}, series = {Proceedings of the 38th ECMS International Conference on Modeling and Simulation (ECMS 2024): June 4th - June 7th, 2024, Cracow, Poland}, booktitle = {Proceedings of the 38th ECMS International Conference on Modeling and Simulation (ECMS 2024): June 4th - June 7th, 2024, Cracow, Poland}, publisher = {ECMS}, isbn = {978-3-937436-84-5}, issn = {2522-2422}, doi = {10.7148/2024-0338}, pages = {338 -- 344}, abstract = {Automated Guided Vehicles (AGVs) and Autonomous Mobile Robots (AMRs) are being applied more and more frequently in a wide range of use cases. Following the general trend of decentralisation of control structure in the Industry 4.0 paradigm, this paper analyses the application of auction algorithms to solve the robot task assignment problem. We focus on the use of mobile robots in production environments with a high level of uncertainty which places hight demand on the flexibility of the online scheduling architecture. A sequential single-item (SSI) auction algorithm is validated for a real-world use case where multiple mobile robots supply car bodies to manual rework stations in a paint shop application using simulation. The optimization objective of such an algorithm is discussed in regards to the examined use case.}, language = {en} }