7. Sustainability in Manufacturing Transforming

The following section illustrates the successful implementation of sustainable manufacturing by comparing three use cases from BMW’s iFACTORY, each with an equal focus on each of the three dimensions but covering the different groups of research objects. With the iFACTORY, BMW addresses the three pillars—LEAN, GREEN, and DIGITAL—thereby setting the direction for the transformation of manufacturing expertise throughout the entire production network (see BMW AG, 2022). This means:

The first use case shows that incorporating innovative circular materials and systems helps to conserve resources and creates ergonomic benefits for associates. To conserve even more resources, the BMW Group has implemented various projects in packaging logistics. These aim to reduce carbon dioxide (CO₂) emissions in cooperation with suppliers and to implement the principles of circular economy to the greatest extent possible. European plants are increasingly using recycled materials for packaging. In 2022, new contracts for reusable packaging in logistics specified almost double the quota of recycled material, increasing from approximately 20% to over 35%. CO₂ emissions are also being reduced through the use of alternative sustainable materials, less single-use packaging, lightweight packaging, and reduced transport volumes. The BMW Group plans and monitors the effects of individual measures via a CO₂ calculator for packaging.

A second example of innovative production processes with positive reductions in energy and water consumption is the so-called dry scrubber. In a major step toward greater sustainability, paint shops no longer wash away excess paint particles with wet scrubbing but instead are switching to a system of dry separation. In the spray booth, any overspray that does not land on the car body is now collected using limestone powder rather than water, thereby considerably reducing water consumption. Another major advantage is that, unlike wet scrubbing, dry separation can be carried out in up to 90% recirculated air. This means that only 10%, rather than 100%, of the air has to be brought up to the required temperature and humidity, thereby saving vast amounts of energy. The limestone powder also does not need to be processed and disposed of, unlike contaminated water. Instead, it can be returned to the material cycle—for use in the cement industry, for example.

The third use case pays in directly to all three dimensions of sustainable production. A 3D human simulation introduces a virtual model of a human into a virtual production environment. It uses a combination of connected planning data to simulate the complete production and assembly process in 3D. Through this, valuable information can be gathered by simple means, such as planned time analysis, ergonomics assessments, workplace optimization, and validation of planning. This enables optimization of process engineering, the conditions for production workers, and process maturity right at the start of production.

Sustainable manufacturing offers a broad spectrum of research opportunities. Due to the interdisciplinary character of sustainability studies, research on the social, economic, and ecological dimensions requires different research competencies. Because of this complexity, this section focuses primarily on the engineering perspectives involving energy, circular processes, and manufacturing technologies and strategies.

With regard to energy in the context of sustainable manufacturing, four main research perspectives can be identified. Improving energy efficiency has long been a major focus of research and practice in the past. In addition to energy efficiency (i.e., the relationship between the value created and the energy used; DIN, 2011), energy flexibility requires consideration in the future (Popp, 2020). Energy flexibility describes the ability of a factory or a process to adapt to a volatile energy supply with no negative effects on productivity, quality, or delivery service (VDI, 2020). Overall, 16 flexibility measures have been identified that can be assigned to the factory, production, or process levels. From a research perspective, manufacturing processes, operations management practices, and digitalization technologies all need to evolve to address both energy flexibility and efficiency.

The second perspective involves the substitution of fossil energy sources with renewable energy sources and technologies within a factory. Currently, a strong trend is evident toward the electrification of industrial processes (Wei et al., 2019). With the decreasing price level of solar panels and increasing battery storage capacity, the integration of volatile energy sources to operate industrial processes with a continuous demand is becoming both feasible and advantageous. Although industrial processes cover a wide range of temperatures, electric heating systems, high-temperature heat pumps, or solar thermal technologies can easily generate lower temperatures up to 140 °C.

The third perspective focuses on the systematic change observed across the entire energy supply chain for electricity, from generation to consumption. Decentralized energy generation using photovoltaic systems can now partially replace the traditional external energy supply generated by large power plants and transported over long distances. These approaches can help reduce costs and increase energy resilience.

Finally, production systems and factories based on direct current represent a major new area of research. These systems allow an easier integration of renewable energy sources, such as photovoltaics, while also eliminating the need for frequency inverters that lead to efficiency losses, such as harmonics, and enabling an easier recuperation of electrical energy (Sauer, 2020). This broad scope of the entire system of energy supply, transport, and consumption reveals tremendous improvement potential for energy efficiency, flexibility, and substitution.

With regard to circular processes, the second area of research in sustainable manufacturing places a strong emphasis on material flows and digitalization. The linear manufacturing approach of “take–make–use–dispose” not only exceeds the waste-carrying capacity of the earth, but has significantly increased the rate of resource extraction in the recent past. In the EU-28, the manufacturing sector generated 10.3% of all waste, making it the third largest contributor after construction and mining (Rashid et al., 2020). Decoupling resource consumption and waste generation from economic growth will require the application of circular manufacturing. The aim of conventional circular or closed-loop systems is to minimize energy and resource inputs, maximize the value generated, and reduce waste and emissions (Nasr & Thurston, 2006). Closing the loop between output and (re)input can be achieved through reuse, remanufacturing, or recycling. In many cases, this approach is limited because the present-day processes and products were not intentionally designed for closed-loop systems, and the effort to implement circularity exceeds the potential benefits.

According to Rashid et al. (2020) and in line with the circular economy definition of the Ellen MacArthur Foundation (2013), a circular manufacturing system is “a system that is designed intentionally for closing the loop of components or products, preferably in their original form, through multiple life cycles” (Rashid et al., 2020, p. 355). Circular manufacturing can operate at the macro-level (e.g., region and smart city), the meso-level (e.g., industrial parks and factory), or the micro-level (e.g., products and processes) (Urbinati et al., 2020). The micro-level is characterized by the shortest loops and thus has the greatest potential environmental benefits. Based on the original 3R concept (reduce, reuse, and recycle), the 6R framework for implementing circular manufacturing systems, which covers the entire product life cycle (reduce, reuse, recycle, recover, redesign, and remanufacture), represents the state of the art for research and practice (Jawahir & Bradley, 2016).

The first R (reduce) refers to the reduction of resource usage in the premanufacturing phase, the reduction of energy and material consumption in the manufacturing phase, and the minimization of emissions in the use phase. The second R (reuse) refers to the multiple life cycles of the original product or its components after each end of life (EOL). The third R (recycle) converts material that would normally be considered waste into new material and process input. To gather the product after the use phase, the fourth R (recover) has the task of recovering the products after their EOL. The fifth R (redesign) incorporates products or components from previous life cycles into the next design concept, while the final R (remanufacture) aims to restore used products to their original state. The 6R system combines traditional methods or tools, such as those for energy efficiency, with innovative remanufacturing processes and facilitates stepwise implementation (Brunoe et al., 2019).

Although circular manufacturing offers tremendous potential for sustainability, its implementation is often hindered by heterogeneous barriers. Because different stakeholders are involved, typically including at least suppliers, the manufacturer, users, and remanufacturing experts, the sharing of data and information is a major challenge. Digital twins of material flows can be used to provide and manage complex and heterogeneous data in discrete manufacturing between them (Acerbi et al., 2022). As an alternative to hierarchical data models, blockchain technology has been implemented to share data among different stakeholders (Govindan, 2022). In doing so, these data models describe the relationships between processes and material flows, reveal optimization potential for circular manufacturing, and deliver consistent and trustworthy data. Thus, in addition to the 6R methodology, the sharing of data and information is considered a prerequisite for implementing circular manufacturing.

Finally, with regard to sustainability in operations, manufacturing technologies and strategies represent a third area of research. On the one hand, innovative processes, such as additive manufacturing (AM) or digitalization technologies, have a strong impact on well-established process chains. On the other hand, further development is required to bring innovative technologies to similar quality levels and process capabilities or to scale them up for manufacturing in batch sizes of single products and high-volume production. On the technological side, additive manufacturing (AM) is a primary area of research. For production scenarios with high complexity and low volumes, AM has already demonstrated competitiveness compared with subtractive or formative technologies (Pereira et al., 2019). Due to the reduction in resource consumption and waste generation, AM has a strong positive impact on sustainability. The main challenge for future AM processes and machines is their integration into complete supply chains that meet the requirements of high complexity and large volumes. Other technological challenges arise during the production of electric cars, particularly battery production, or the production of components for hydrogen applications. Both of these examples require innovative, isolated process steps, as well as completely new entire production systems and machines; consequently, low quality levels with high fluctuations are a major concern and have a negative impact on overall equipment effectiveness (OEE) (Schnell & Reinhart, 2016). Finally, process chains for innovative applications or AM will not replace traditional technologies. Further potential for improvement lies in the adoption of hybrid manufacturing approaches, such as configuring the most suitable manufacturing technology for a best practice process chain or even combining technologies with the machine tool (Merklein et al., 2016).

Digitalization and the use of artificial intelligence offer future research perspectives regarding sustainability. At present, Industry 4.0 approaches have been used primarily to address the environmental dimension, but researchers have already outlined research agendas to address the social and economic dimensions in a holistic approach (Machado et al., 2020; Stock & Seliger, 2016). Digitalization techniques, such as the Internet-of-things (IoT) or cloud manufacturing, represent technological tools that must be adopted to pursue sustainability objectives. Artificial intelligence (AI) can be used to manage the complexity of sustainability-related data (e.g., with big data analytics approaches). In any case, digitalization and AI require access to reliable data at the process level.

Manufacturing strategies are an additional area of research. Due to the cross-dimensional nature of sustainability, its strategy must be strongly linked to functional strategies, such as product or process development. Sustainable manufacturing involves technological aspects as well as methods and tools; therefore, a challenge for future research is to integrate well-established management processes, such as quality and supply chain management, and production systems, such as lean management, with sustainability approaches. Replacing existing processes and tools is not recommended; rather, these should be further developed by considering sustainability aspects (Pampanelli et al., 2014).

In summary, various aspects of future research on energy, circular processes, and manufacturing technologies have been highlighted, without claiming to be exhaustive. An important point to note is that intrinsically motivated employees drive the transformation to sustainable manufacturing. They use valid and real-time data in their decision-making to achieve specific and individual sustainability goals. Therefore, in addition to the technical and organizational challenges described above, a suitable qualification concept is of particular importance. To achieve broad acceptance for the implementation of sustainable manufacturing, specific training content and programs with theoretical and practical content must be developed for all hierarchical levels within a company.

Research and practice agree that sustainability aspects can only be addressed if improvements consider all phases of the product life cycle (i.e., the development phase, the manufacturing phase, the use phase, and the end-of-life phase). Interactions between the phases need to be considered (Liu et al., 2019). In the following section, challenges and opportunities along the design and manufacturing life cycle phases that impact sustainability will be discussed.

The design of a manufacturing system is critical because changes in later stages can only be implemented with a substantial effort. Moreover, decisions have to be made under uncertainty and undefined boundary conditions. The dimensioning of a production system, and especially its capacity, is closely related to its sustainability impact and must therefore be derived using a systematic process. For example, if the technological and production capacity after ramp-up significantly exceeds the current process demand, this effect leads to inefficient operating points in the manufacturing phase. In addition to sizing, the specification of the production equipment in terms of its process steps and manufacturing technology is of great importance. Value-adding steps must be optimized, while non-value-adding steps and process waste must be minimized. If we are to manage the sizing uncertainty and define optimal processes, we must have reliable process data. However, collecting process data based on physical testing and design of experiments (DoE) is not only time-consuming; it is also often impossible to obtain because the manufacturing equipment is not yet available at this early stage. As digital twins and simulation technologies provide virtual representations of systems along the life cycle, they can also be used to model the dynamic behavior of the production system in terms of sustainability (Negria et al., 2017). Input factors, such as raw materials, consumables, and energy, and output factors, such as productivity or waste streams, can be determined based on varying operating conditions without performing physical tests. To minimize the implementation effort and to achieve high accuracy of digital twin modeling, most practitioners and researchers follow a systems engineering approach (i.e., the production system is broken down into smaller units for which reliable digital twins are developed based on existing data; Computer-Aided Design [CAD], Product Lifecycle Management [PLM], etc.). With increasing maturity and given the physical availability of manufacturing equipment and process design, congruence between digital twins and physical systems must be achieved. Finally, the increasing application of digital twins not only increases the efficiency in advanced product quality processes (APQP) within a single company, but they can also be used to model interdependencies related to sustainability at the interorganizational level.

During the ramp-up and in series production, the focus must be on efficiency. One key metric is overall equipment effectiveness (OEE), with its three components: loss of availability, loss of performance, and loss of quality (Focke & Steinbeck, 2018). Since losses of availability include all downtime of the manufacturing system, the indicator shows the percentage of a period that the system is in stable operation. Sustainability is negatively impacted because a high level of availability losses requires additional capacity reserves to meet the total demand, with a negative impact on space, and frequent interruptions to operations result in ramp-up losses of energy, personnel, and raw materials. The second component of OEE is performance loss, which describes whether the production system is at its optimal operating point regarding energy consumption, material input and output, and personnel. Temporary or permanent deviations require additional production capacity. Finally, quality loss is the amount of scrap and rework that occurs in the process chain. Poor quality levels have a direct impact on sustainability because the initial raw material is not processed into finished goods. Emissions, material consumption, etc. from the raw material generation phase have already been incurred but cannot be used to create products or added value. To incorporate sustainability aspects, a stronger focus is needed on inputs and outputs that have not been considered before, such as emissions and waste streams. These extensions to existing key performance indicators will provide reliable and consistent data for effective sustainability decision-making. Practice and research reflect that modern shop floor management systems follow this approach and include sustainability metrics. Based on this information, anomalies from the defined operating points can be identified, and appropriate countermeasures can be initiated on the shop floor (Cerdas et al., 2017).

For the production system design and manufacturing phases, efficiency is the central objective. Industrial production processes transform material, energy, and other inputs into finished goods, delivering added value as well as by-products, such as waste streams and energy losses. To minimize these by-products, circular production processes must be developed and installed (Gupta et al., 2021). The concept of the ultra-efficient factory relies on reuse of all types of waste and energy in two main energy and material recycling loops. In the first loop, wasted energy and materials are fed directly back into the manufacturing phase. In the second loop, the product is returned to the supply chain at the end of its useful life. This minimizes downcycling of material (i.e., the use of material for lower performance applications). This means that sustainability is based on efficient product generation processes and on efficient recycling and remanufacturing concepts that must be designed into the manufacturing design phase.

Digitization, Industry 4.0/IoT, and artificial intelligence have significant potential to allow manufacturing companies to implement sustainability (Stock & Seliger, 2016). However, an important point to consider is that digitization is not an end in itself, and its implementation requires a systematic approach. A generic model is the manufacturing analytics approach with four levels: (1) visibility, (2) transparency, (3) forecasting ability, and (4) prescription. This approach has been developed for the systematic implementation of digitization technologies (Meister et al., 2019). It ranges from lower levels of digitalization, such as simple data collection, to the modeling of complex system behavior using artificial intelligence. Although it is not primarily intended for the implementation of sustainability, it represents a systematic approach to the acquisition, handling, and management of data for specific objectives. When applied to sustainability issues, this analytics approach can be used to better understand correlations, optimize processes, and anticipate and prevent negative impacts on the three dimensions of sustainability.

The objective of first-level visibility is to capture data from the shop floor. Accessing shop floor data is a hurdle because it either has to be accessed through a wide variety of different protocols (OPC Unified Architecture [OPC UA], MTConnect, EuroMap77, etc.) or is not accessible at all. Since sustainability data are not always part of existing protocols, retrofitting existing machines with Internet-of-things–compatible sensors is often necessary. The result is that the aggregated data at each time step are stored on a common, often cloud-based platform. The objective of transparency (2) is to systematically identify the root causes of specific problems and deviations, such as the increased use of energy or material consumption. Individual and specific KPIs, for example, for different functional units, can be extracted. The forecasting ability (3) enables us to make projections of trends in the future and to proactively manage deviations. This ability can be used, for example, for the demand-side management of production equipment and the ramp-down of lower priority processes and machines in the event of energy shortages or price increases. At the prescription level, courses of action are being proposed.

As described above, the analytics approach and the application of artificial intelligence methods and tools are highly interdependent. AI algorithms require consistent data on a continuous basis, which is provided by the four-stage model (Weber et al., 2019). When this condition is met, AI methods can first be applied to reduce the complexity of data lakes. For example, principal component analysis can be used to identify the primary drivers of sustainability improvement actions. In addition, black-box AI systems, such as neural networks, can be applied to speed up simulation runs of energy consumption under different or uncertain conditions.

The development of methods and standards for sustainable production has long been a focus of research and practice. The goal is to make visible the relationships between production, consumption, and disposal and to assess the impacts of economic activities. Life cycle assessment (LCA) has emerged as the most important and accepted method from a technical perspective (Hagen et al., 2020). It is embedded in the ISO 14000 series of environmental standards that address environmental management issues associated with production processes and services. Common to all sustainability standards is the breaking down of industrial value streams into process modules for which mass flows (raw materials and fuel inputs, products, by-products, and waste), energy inputs, and emissions to water, air, and soil are analyzed. While the data of Scopes 1 and 2 of DIN EN ISO 14064 can be collected internally, cooperation with suppliers is required to collect data for Scope 3 raw and operating material inputs. Scope 3 CO₂ emissions are particularly relevant, as they can account for up to 50% of the total footprint (Gross & Hanenkamp, 2021). In practice, suppliers are under increasing pressure from their customers to provide data on CO₂ emissions.

Environmental impact categories are assigned to the life cycle inventory analysis, and their quantification allows us to focus on prioritized environmental impacts. In practice, the application of sustainability standards with precise data requires a high level of technical effort due to its complexity, as well as extensive methodological knowledge and expertise. As a result, assessments are often conducted on a project-by-project basis and are static in nature, making them unsuitable for the operational optimization of production processes. A dynamization of the LCA (i.e., continuous generation with real-time data) can be used to derive precise measures for the operational optimization of the production processes on the shop floor (Cerdas et al., 2017). Finally, to reduce the burden on all stakeholders in the supply chain, the exchange of sustainability-related data based on trust and using reference data models is required.

The need for more efficient and sustainable operations necessitates that we do not develop and optimize production systems independently, but rather consider them as interconnected entities. In the circular concept, energy, material, and waste streams from one process must be considered for secondary use in other processes. This can only be achieved by coupling different entities of the manufacturing system. The concept of system coupling allows the physical flow of materials between subsystems, the recuperation and use of wasted energy, such as electricity or heat, and the exchange of information, such as future demand or the current status. The peripheral components within production systems, such as cooling devices, are typically operated using simple control strategies with few set points; consequently, energy demand peaks cannot be avoided. Heating, ventilation, and air-conditioning systems (HVAC) rely on more complex control strategies, but their control parameters are not adapted to upcoming heating or ventilation demands. The prerequisite for system coupling within the factory is the exchange of data between manufacturing systems and technical building equipment. This allows for the identification of optimal operating points that lead to the adjusted control parameters of the subsystems.

Beyond the internal system coupling within the factory boundaries, the manufacturing site also interacts with the local urban environment. Historically, manufacturing and urban spaces have coexisted, and negative impacts have led to the location of factories on the outskirts of cities. Urbanization, as a megatrend, forces the development of new concepts, such as urban manufacturing or the integration of manufacturing in smart cities (Matt et al., 2020). By definition, an urban factory is not only a factory that is simply physically located in an urban environment; it is one that strongly interacts with other urban entities regarding information, material, and energy flows and that relies on the local market and suppliers (Ijassi et al., 2022). In this way, urban production can contribute to the sustainable development goals (SDG) of affordable and clean energy (SDG 7), decent work and growth (SDG 8), industry innovation and infrastructure (SDG 9), sustainable cities (SDG 11), and responsible consumption and production (SDG 12) (Juraschek et al., 2018). Thus, negative impacts, such as emissions, arise, but positive contributions, such as the availability of jobs in urban production scenarios, also occur. Given the global trend of urbanization, smart cities will also play a central role in sustainability. Although the concept of a smart city has no common definition in research and practice, it has a broader scope than manufacturing (Suvarna et al., 2020). It encompasses all entities within the city (including buildings, transportation, energy grids, health care, manufacturing, and commercial services) that need to be connected. This also means that material, information, energy, and people flows need to be considered and optimized in the context of the city ecosystem.

Four different principles of the sustainable factory of the future were discussed in this chapter. First, the planning and operation of manufacturing processes with respect to sustainability were shown. While, in the planning phase, the dimensioning and specification of the production system are crucial, whereas, in the manufacturing phase, the focus has to be on efficiency and abnormality management. In the future, this will require, for example, bringing sustainability aspects to the daily shop floor management level. Second, the availability of real-time process and manufacturing data is critical. Sustainability-related decisions are often highly complex and require a systematic approach to collecting, processing, and proactively applying manufacturing data. Third, the factory of the future can be assessed for sustainability based on accepted standards. Compared with today’s static nature of assessments, the assessments will need to be performed more frequently with minimal effort. Finally, the sustainable factory of the future is characterized by system coupling at multiple levels. Within the factory, production systems, peripheral components, and HVAC systems are physically and digitally connected and operated with global optima in mind. Beyond the physical boundaries of the factory, the exchange of material, information, and energy with the urban space must be considered to have a positive impact on sustainability. Given these directions, research and practice are challenged to develop methods and tools for implementing sustainable factories.