autoregion international - Edition 1/2020

33 The following measures are available for this [5] : ›  Scalable assembly modules linked by flexible automated guided vehicles (AGVs) and elimination of permanently installed conveyor ›  Increased use of universal industrial ro- bots in cooperation with production and assembly workers ›  Utilisation of the potential offered by digitalisation and the increasing perfor- mance of information technology for au- tomotive production (see Fig. 1) In various publications [6] , management consultancy strategy& has demonstrated methods for automakers; one of these – which we will elaborate on below – is that of Flex Champions (Fig. 2): vehicle production according to this business model pursues the approach of maximum adaptability. Various models, drive variants and deriv- atives with a large spread of work content can be cost-effectively produced if each body takes its own individual path through the assembly modules, which in turn are supplied on time with components from pre-selected shopping baskets by AGVs. This scenario requires permanent locali- sation and online tracking of body shells, AGVs for their transport, assembly parts and their means of transport. Conveyor technology across several levels such as in today’s automotive plants is then no longer necessary; even track-bound ve- hicles will be largely eliminated from the factory. Today’s (mostly outsourced) lo- gistics with external supplier warehouses, supermarkets, milk-run trains to supply the assembly line and Kanban racks at the conveyor belt are changing toward picking sets of components for particular installa- tion volumes. Only C-parts will be provided just-in-time in the modules (managed ex- ternally). Simulation and operational control us- ing new methods such as reinforcement learning for all autonomous vehicles (see example in Fig. 4) which transport the workpieces and deliver the components is challenging. It has not yet been proven that such a novel production and control system can be at least as efficient but more adaptable as the sophisticated and synchronised as- sembly line production which has been tried and tested over many years [8] ; nor has it been proven that such a system can au- tonomously and robustly respond to unex- pected events. Self-organisation is taking place not only in the factory but also at the level of global supply chains. Many suppliers find it difficult to hand over planning au- tonomy and confidential information to a central supply chain orchestrator. This results in self-organisation and agent- based decentralised planning tasks. Ini- tial attempts, e.g. production of the e.Go vehicle in Aachen with small numbers for an automotive plant, indicate that there is great potential for self-organisation, even though several outstanding ques- tions remain to be answered. strategy& and Fraunhofer have set themselves the goal of answering these outstand- ing questions and opening up paths to new production systems for their customers. Sources [1] acatech (Pub.): Umsetzungsempfehlungen für das Zu- kunftsprojekt Industrie 4.0 – (Implementation recom- mendations for the Industry 4.0 future project) Final report of the Industry 4.0 work group, April 2013 [2] Sauer, O.; Sutschet, G.: Production Monitoring linked to object identification and tracking – a step towards real time manufacturing in automotive plants. In: Teti, R. (Ed.):Proceedingsofthe5thCIRP InternationalSeminar on Intelligent Computation in Manufacturing Enginee- ring, Ischia (Italy): 2006, pp. 321-326 [3] Bussmann, S.; Schild, K.: Self-Organizing Manufac- turing Control: An Industrial Application of Agent Technology. In Proc. of the 4th Int. Conf. on Multi-agent Systems (ICMAS’2000), Boston, MA, USA, 2000, pp. 87- 94. [4] Weyer, M.; Spath, D.: Das Produktionssteuerungs- konzept “Perlenkette”. Herausforderungen und Hand- lungsempfehlungen der Implementierung. (The “pearl chain” production control concept. Challenges and recommended action for implementation.) ZWF Zeit- schrift für wirtschaftlichen Fabrikbetrieb; 2009, No. 12; pp. 1126-1130. [5] Fraunhofer Allianz Automobilproduktion (Pub.): Mobili- tät der Zukunft muss produziert werden. (Future mobi- lity must be produced.) Position paper, Chemnitz: 2019. [6] Weber, H. et al: Transforming vehicle production by 2030 – how shared mobility and automation will revolu- tionize the auto industry. strategy&White Paper, 2018. [7] Fraunhofer IOSB: Künstliche Intelligenz für die Produk- tion von morgen. (Artificial intelligence for tomorrow’s production.) White paper, Karlsruhe: 2019. [8] Monden, Y.: Toyota Production System: an integrated approach to Just-in-time. Engineering Management Press, 1997. Fig. 3: Example of a first application of self-organization in automobile production Fig. 4: Self-organisation using the example of swarming insects [7] Autonomous Systems  www.iosb.fraunhofer.de/ servlet/is/99090/ Business Unit Automation and Digitalization www.iosb.fraunhofer.de/ servlet/is/12579/