Needing Neither Conveyer nor Cycle – Modular Production in Practice

In the field of electromobility, companies face extremely high demands in terms of flexibility and adaptability of production. MAN Truck & Bus SE therefore relies on a modular production system in its first battery factory to simultaneously meet these requirements and ensure efficient large-scale production. The basis for the dynamic control of such modular production, without a rigid assembly line and uniform work cycle [1], is the end-to-end digitization of processes and an intelligent production control system.

Read the German version of this article here.

A new chapter has also begun for commercial vehicle manufacturers with electrification [2]. At MAN, a commercial vehicle battery consists of two to four levels, so-called layers (Fig. 1), where the battery modules are grouped together with the battery cells they contain, as well as a battery junction box (BJB), which is mounted on the side as a switching unit.

Despite their different designs, the batteries have a consistent energy storage capacity and enable a modular usage concept in which the number of battery systems can be flexibly adapted to individual customer requirements. Four to eight high-voltage batteries are installed in the electric buses and three to six in the new electric trucks. However, the customer-specific battery configuration and the three battery variants, some of which have different production contents and assembly sequences, require a highly flexible production system. The highly dynamic nature of electromobility also requires the production concept to be highly adaptable.

The concept of „modular (matrix) production“ was therefore chosen as the organizational form for large-scale production [3, 4]. This should guarantee the high demands on flexibility and adaptability in addition to the efficiency of large-scale production. On the one hand, these result from the new product, the different product variants, dynamic product development, new production processes and a varying product mix as well as the efficient implementation of different operating points in production. On the other hand, it is necessary to reconfigure the production system with as little effort as possible to realize the gradual increase in production capacities, the integration of further product generations and technological advancements.

Basic principles of a modular production system

The organizational form of modular production is characterized by three central basic principles [3]: decoupled stations, a self-organized production system and an integrated overall system (Fig. 2).

The design of the freely interlinked stations combines the high division of labor of large-scale production with the flexibility and versatility of stand-alone production in that parallel stations sometimes enable longer cycle times (e.g. for setting all modules in a station) as well as the step-by-step assembly and conversion of individual parallel stations.

The self-organized production system is characterized by a dynamic comparison between the requirements and the available capacities. The current production orders are assigned with their respective work content, assembly sequences and processing times to the available production and transport capacities in short cycles based on availability and capacity utilization. In this way, demand and status are constantly compared and rescheduled in the modular production system. The integrated overall system describes the embedding of necessary logistics and quality processes into the planning and control of modular production. A directly adjacent high-bay warehouse for components and finished batteries as well as an automated and largely unmixed material supply enable efficient organization of production and logistics. The required quality and completion processes can be planned directly into the production process thanks to the dynamic demand capacity comparison and avoid downstream completion areas.

Eight features of a modular production system

All eight features that characterize a modular production system are used in the new battery production:

1. Without conveyor belt: Flexible interlinking of the stations by driverless transport vehicles and thus an individual flow of the individual orders based on the respective production requirements and assembly sequence.
2. Without cycle: Considering the variant-dependent processing times of the different product variants in the respective stations between one and fifteen minutes.
3. Self-control of orders: Variable planning (sequencing) and allocation (scheduling) of individual production orders to the stations by an intelligent production control system.
4. Self-control of resources: Variable assignment of one or two employees to each manual station, resulting in different operating points in each shift.
5. Single-variety provision directly at the installation site: Automated and largely single-variety material provision and therefore generally one-off handling immediately before installation.
6. Integrated quality and short control loops: Possibility to react immediately to deviations directly in the station as well as by feeding them out and back into the regular process.
7. Adaptation to employees: Employee-oriented design through digital worker guidance as well as variable target time and number of employees during qualification.
8. Adaptation to changes: Easier expandability and conversion during operation in the event of changes to product variants or the product mix, new product generations or production technologies.

MES with AI component due to high requirements

The implementation of modular production poses major challenges for industrial companies, primarily in the area of IT system architecture. In order to ensure the desired flexibility in terms of quantities, sequences and routes, the IT systems need to be networked horizontally and vertically throughout.

In this architecture, the Manufacturing Execution System (MES) plays a decisive role in the control and monitoring of production processes.

The aforementioned requirements for flexibility and scalability on the one hand and robustness and quality on the other pose major challenges for production planning and control (PPC). Traditional approaches to PPC reach their limits especially here. Therefore, an already established MES is being expanded for this use case with the help of AI-based optimization components. The aim is to continuously react to changes in production in order to ensure effective and efficient production processes. Figure 3 shows the IT architecture including the superordinate and subordinate systems.

Manufacturing software for modular production processes

The MES takes the production orders for the HV batteries from the higher-level ERP system (Fig. 3; A) and firstly puts them into coordinated sequences for layer, BJB and pack assembly (sequencing). The sequenced orders are then broken down into their individual work steps and assigned to the workstations (scheduling). This is made possible by managing the individual work steps and their dependencies, such as predecessor and successor relationships, in the form of a priority graph within the MES.

In addition to scheduling, the retrieval and transport orders are also generated and transmitted to the warehouse management system or the AGV fleet control system. The entire scheduling process takes place cyclically and, if necessary, event-based according to the status and completion messages from all workstations. The MES receives progress feedback on the respective work steps from the worker information system (WIS), which is used in all workstations (Fig. 3; B).

The WIS is responsible in all workstations, among other things, for the shoring check, the monitoring of bolting cases and the continuous transmission of the processing status and the remaining completion time to the MES. The WIS is also responsible for visualizing the assembly steps in the manual stations. Based on this feedback, scheduling that has already been carried out can then be repeated and rescheduled in the event of production deviations, such as delayed processing (Fig. 3; C).

Special features in the production of battery systems

When dynamically assigning the next work step to be planned to the workstation, a large number of different resulting criteria and objectives must be taken into account in addition to existing framework conditions:

1. General conditions

• Production system consisting of automatic stations and manual stations
• Variable cycle times at the workstations with processing times between 1 and 15 minutes
• Up to five parallel stations per station group
• Forward-looking planning and transfer of transport orders to the AGV fleet control system required to optimize resources
• Scheduling of up to 4,500 operations per shift

2. Criteria and objectives

• High adherence to delivery dates
• Minimizing an order’s throughput time
• Minimization of work in progress
• High and even utilization of the workstations
• Prioritization of workstations per task type
• Minimization of transport and waiting times

Qualicision AI is used in the MES to solve this multi-criterion decision-making task. Qualicision AI’s modeling of the decision task is intuitive and flexible, allowing the production control center to understand the effect of criteria in relation to the target values. Specific training courses are held to qualify employees in the use of the AI component so that fine-tuning the targets is possible at any time. Furthermore, additional work steps or workstations as well as additional criteria can be integrated into sequencing and scheduling without the need for time-consuming reprogramming of the algorithms. This is crucial in terms of the high demands placed on the flexibility and versatility of battery production.

Efficiency and effectiveness thanks to modular production

Modular production can present the answer to effectively and efficiently manufacturing HV batteries in volume production despite product diversity and short development cycles. This results in significant advantages, such as more flexible handling of product changes and production adjustments. At the same time, modular production requires a smooth flow of information across all IT systems and appropriate employee training. Planning and controlling this assembly organization without using the potential of AI seems almost impossible due to its complexity. This requires an intelligent MES in a holistically coordinated IT architecture.

In summary, flexibility and adaptability of production in the field of HV batteries are key requirements that can be met by the concept of modular production despite the existing challenges.

Open Access DOI: 10.30844/FIE.24.4.WK

Das könnte Sie auch interessieren

Agil planen, flexibel fertigen Effiziente Prozesse bei hoher Variantenvielfalt

In der Serienmontage bestimmt die Austaktung, welcher Arbeitsvorgang an welchem Arbeitsplatz erfolgt. Doch moderne Montagelinien werden immer komplexer: Mehrere Modelle und zahlreiche Varianten laufen parallel auf derselben Linie. Dadurch steigt die Zahl der Arbeitsvorgänge und möglichen Zuordnungen stark an. Schon kleine Änderungen bei Variantenmix, Arbeitsinhalten oder Taktzeiten können die Belastung einzelner Stationen deutlich verschieben.

„KI ersetzt keine Erfahrung – sie macht sie skalierbar!“ Im Gespräch mit Jürgen Schön, Head of Manufacturing Industry GTM EMEA bei ServiceNow

Die Fertigungsindustrie hat große Fortschritte bei der Einführung von KI gemacht – doch meist bleibt es bei Pilotprojekten. Entscheidend wird nun, KI ganzheitlich entlang der Wertschöpfungskette zu verankern, um echte Wirkung zu erzielen: höhere Produktivität, stabilere Prozesse und weniger Ausschuss. Immer mehr Unternehmen entwickeln konkrete Anwendungsfälle und setzen diese um. KI wird damit vom Experiment zum festen Bestandteil operativer Exzellenz.

Innovation, Digitalisierung und KI im Maschinenbau Im Gespräch mit Prof. Dr. Oliver Amft, Institutsleiter bei Hahn-Schickard

Hahn-Schickard und das IMIT Villingen sind Innovationsmotoren für Mikrosystemtechnik und Digitalisierung. Mit Hightech-Lösungen in Bereichen wie Sensorik, Embedded Systems, KI und digitaler Medizin unterstützen sie Unternehmen – vom Start-up bis zum Großkonzern – entlang der gesamten Wertschöpfungskette. Forschung trifft Praxis – regional verankert, international vernetzt.

Making Manufacturing Easy: beständig, innovativ, spezialisiert Im Gespräch mit Dr. Clemens Pieper, Geschäftsführer Deutschland bei Monitor ERP

Langfristig orientiert und gleichzeitig Innovativ – seit über 50 Jahren treibt Monitor ERP die Digitalisierung im produzierenden Mittelstand voran. Mit Cloud, KI, Nachhaltigkeitsfunktionen und tiefem Branchenwissen bietet das System eine zukunftssichere ERP-Lösung, die schnell implementierbar, erweiterbar und praxisnah ist – ausgezeichnet mit dem Factory Innovation Award.

Retrofit in der Produktionslogistik 5 Punkte, die Sie unbedingt beachten sollten

Das Retrofit bestehender Produktions- und Logistiksysteme bietet Unternehmen die Chance, mit überschaubarem Aufwand ihre Effizienz zu steigern und Kosten nachhaltig zu senken. Trotz klarer Vorteile zögern viele Betriebe, da sie wirtschaftliche Unsicherheit und Betriebsunterbrechungen fürchten. Dabei können bereits kleine Modernisierungen schnelle Erfolge bringen und gleichzeitig die Weichen für künftige Herausforderungen stellen.

Factory Cloud und Datenräume – Zukunft der Produktion Flexible, sichere und datengetriebene Fertigung

Produzierende Unternehmen stehen unter wachsendem Druck, ihre Prozesse flexibel und effizient zu gestalten. Die Factory Cloud ermöglicht die lokale Verarbeitung von Produktionsdaten, während Datenräume den sicheren Austausch sensibler Informationen entlang der Wertschöpfungskette gewährleisten. Gemeinsam bieten sie die Grundlage für eine souveräne, datengetriebene Produktion.