Integrative computational design and construction: Rethinking architecture digitally

Increasing the construction capacity, while at the same time significantly reducing harmful emissions and consumption of nonrenewable resources, and still providing a liveable and affordable built environment, provides a great challenge for future construction. In order to achieve this, both the productivity of construction processes and the energy and resource efficiency of construction systems have to be improved in a reciprocal process. Digital technologies make it possible to address these challenges in novel ways. The vision of this Cluster of Excellence IntCDC at the University of Stuttgart and the Max Planck Institute for Intelligent Systems is to harness the full potential of digital technologies to rethink design and construction based on integration and interdisciplinarity, with the goal of laying the methodological foundations to profoundly modernize the design and construction process and related building systems by adopting a systematic, holistic and integrative computational approach. One key objective is to develop an overarching methodology of “ co-designing ” methods, processes and systems based on interdisciplinary research encompassing architecture, structural engineering, building physics, engineering geodesy, manufacturing and systems engineering, computer science and robotics, and humanities and social sciences. In this way, the Cluster aims to address the ecological, economic and social challenges and to provide the prerequisites for a high-quality and sustainable built environment and a digital building culture.

and the energy and resource efficiency of construction systems have to be improved in a reciprocal process. Digital technologies make it possible to address these challenges in novel ways. The vision of this Cluster of Excellence IntCDC at the University of Stuttgart and the Max Planck Institute for Intelligent Systems is to harness the full potential of digital technologies to rethink design and construction based on integration and interdisciplinarity, with the goal of laying the methodological foundations to profoundly modernize the design and construction process and related building systems by adopting a systematic, holistic and integrative computational approach.
One key objective is to develop an overarching methodology of "co-designing" methods, processes and systems based on interdisciplinary research encompassing architecture, structural engineering, building physics, engineering geodesy, manufacturing and systems engineering, computer science and robotics, and humanities and social sciences. In this way, the Cluster aims to address the ecological, economic and social challenges and to provide the prerequisites for a high-quality and sustainable built environment and a digital building culture.

K E Y W O R D S
co-design, digital building culture, digital technologies, integrative computational design and construction, interdisciplinary research The challenge for building in the future is to build more, while emitting fewer pollutants and using fewer nonrenewable resources, and still create a high-quality and liveable built environment. In a mutually influencing process, both the productivity of the building process and the energy and resource efficiency of the building systems must be the Max Planck Institute for Intelligent Systems is to use the full potential of digital technologies to rethink planning and building in an integrative and interdisciplinary approach and thereby laying the methodological foundations for a comprehensive modernisation of building design. A central objective is the development of an overarching methodology of "co-design" of methods, processes and systems, based on interdisciplinary research between the fields of architecture, civil engineering, engineering geodesy, production and systems engineering, computer science and robotics, as well as the humanities and social sciences. The aim is to identify solutions to the ecological, economic and social challenges and to create the conditions for a high-quality, liveable, and sustainable built environment as well as for a digital building culture.
In order to limit man-made global warming to 2 C, as agreed in the 2015 UN Climate Change Conference in Paris, we must drastically reduce greenhouse gas emissions. Unlike mobility or industrial production, climate-relevant improvements in buildings can be achieved quickly and effectively. Research and development have therefore focused intensively on the energy optimisation of building operation in recent decades. Due to the continuously tightened legal framework conditions, a nearly climate-neutral building stock by the year 2050 seems realistic from today's perspective. The focus is thus shifting from the operation of the buildings to the construction and disposal of the building structures and the CO 2 emissions and gray energy generated in the process. 1 In addition, there is a high demand for additional buildings. In 2050, the world population will be almost 10 billion people. To provide dignified housing for all, a living space roughly equivalent to today's global building stock will be needed over the next three decades. 2 This enormous need does not only arise in Africa and Asia. In Germany, too, 400 000 housing units per year need to be built, mainly due to demographic change. 3 Currently, not even half of this demand is being met.
The productivity of the construction processes must therefore be increased considerably, but in fact it has been stagnating for decades. 4 The climate crisis and population growth pose unprecedented challenges for buildings. Productivity and resource efficiency must be reconciled with architectural diversity, because without cultural and social acceptance, the opportunities of industrialized building cannot be exploited. This will only succeed with leaps in technology, as we are seeing in other areas, such as electromobility or the energy transition.
Instead, research and development in construction is all too often limited to incremental improvements of existing methods and processes.
Compared to all other industries, construction has the lowest level of digitalisation, even far behind agriculture and forestry. 5 This is reflected, among other things, in the stagnating or even declining productivity of the construction sector for decades and the associated increase in the cost of construction services. Efforts to increase digitisation in the construction sector are currently focused primarily on building information modeling (BIM), which concentrates on increasing the efficiency and reliability of existing processes in the planning and operation of buildings, while production processes in prefabrication or on the construction site are not yet addressed. 6 In addition, there is a steadily growing number of initiatives that call for and promote increased digitalisation of construction. However, the small-scale structure of the construction industry and a correspondingly fragmented research landscape favors incremental approaches that mostly follow a similar basic idea of digitalisation: Digital technologies are primarily used for the "computerisation" of existing processes and procedures in order to achieve a higher degree of automation of prefabrication and site-based construction. 7 This, too, will only be a transitional step-automated building processes for predigital construction methods and building systems are created, but the full potential of digital technologies for integration and innovation cannot be raised. In order to fully exploit the possibilities of digitalisation, fundamentally new approaches for sustainable construction are necessary. It is expected that the comprehensive interdisciplinary methodological findings and research results of the Cluster will lead to integrative approaches to the use of digital technologies that will help to address the environmental, economic and social challenges that cannot be solved with current incremental approaches. In this way, the foundations will be laid for a high-quality, liveable and sustainable built environment and a digital building culture in the future. The research objective is therefore to develop biaxial spanning timber building systems for multi-story construction, long-span, fiber composite load-bearing structures, components made of graded concrete and façade elements made of biomaterials that use digital planning and fabrication to reconcile architectural diversity, constructive performance and resource efficiency.
Furthermore, architecture, in which each building is typically unique, can become a model for integrative and individualized design and production processes in other areas. It can be assumed that the expected findings will also be highly relevant beyond the building sector.

| CO-DESIGN AS AN OVERARCHING RESEARCH APPROACH AND METHODOLOGY
To unlock the full potential of digital technologies for future construction, the Cluster of Excellence follows an overarching research approach of "co-design" of the following research areas: A-planning and engineering methods, B-manufacturing and construction processes, C-material and building systems. A further research area D considers building culture, regulatory, social and ecological requirements and draws on historical experience with earlier approaches to digitalisation, modularisation and automation. In this area, for example, a holistic quality model is being developed to ensure the technical, ecological and social quality of co-design processes and products. 8 Likewise, architectural-historical cross-references to the approaches of industrial and serial building of the 1960s and 1970s are examined and reflected upon to enable learning from history from the very beginning. E-building demonstrators will be used to test the architectural integration and validate the research results.
A research culture that examines the above-mentioned research areas A to D separately can be seen as the main barrier to innovation in the construction sector. Thus, we observe that digitisation in current research usually seems to be anchored either in the area of computer-aided design and construction methods, in the area of developing digital manufacturing and construction processes, or in the area of improving material and building systems. Consequently, research results are limited to one area and mostly work incrementally. 5 Digital manufacturing for example, is being developed with a focus on automated production of existing building systems using known design and construction methods. Similarly, database-driven design (eg, BIM) is developed based on datasets of known, predigital building systems and associated established construction processes.
To realize the full potential of digitalisation, it is essential to adopt a more integrative approach that conducts research in all these areas simultaneously and in a feedback loop.
Such comprehensive innovation cannot be expected from the construction industry itself due to the current business models and associated narrow perspectives of the various stakeholders in the sector. In order to advance the construction sector in a profound and comprehensive way, large-scale basic research is required to develop methods that enable future-proof innovations.
The research in the Cluster aims to establish an overarching methodology of co-design of design and analysis methods, manufacturing and construction processes as well as material and building systems ( Figure 1). In this context, co-design is also understood as co-optimisation. Our definition of co-design is an integrative approach that transcends disciplinary boundaries and silos, Co-design is based on the simultaneous and feedback driven development of generative design methods for exploration, analytical methods for optimisation, monitoring methods to capture actual behavior, cyberphysical processes of digital prefabrication and robotic construction on the building site, and also considers multifaceted stakeholder perspective. As a result, new building systems are being created for the most important architectural applications, that is, multi-story and long-span buildings, building envelopes and the densification of the existing building stock, considering social needs and expectations, environmental impacts, regulatory requirements and historical experience.
Specifically, work is currently underway on long-and two-axis spanning, multi-story timber building systems, long-span, high-performance, lightweight roof support structures made of fiber composites, as well as components made of graded concrete and façade elements made of biomaterials. In the later project phases, the focus will then be expanded to include the extension of existing building stock. Therefore, our research approach is integrative, considering multiple disciplines and phases simultaneously, and inductive, developing a general methodology from the study of specific cases representative F I G U R E 1 Co-Design of A, planning and engineering methods, B, manufacturing and construction processes, C, material and 30 building systems, D, cross-cutting social, environmental and humanities issues, and E, building demonstrators. Source: IntCDC University of Stuttgart of important architectural applications. In contrast to the prevailing sequential integration in the form of a digital chain, the Cluster explores co-design as a methodology for the simultaneous and feedback integration of geometric, structural, mechanical, hydrothermal, acoustic properties, environmental, economic and social factors, esthetic and spatial qualities, and possibilities of cyber-physical manufacturing and construction processes.
Current design approaches assume that the exact location, dimension, specification and quality of each building element can be precisely defined and then serve as a clear blueprint for construction.
But already today, many construction projects have reached a level of complexity that makes it increasingly difficult to plan, manage and execute them in such a completely deterministic way. Co-design not only enables a much higher level of integration in the planning and development phase; the linking of methods, processes and systems also lays the foundation for behavior-based, sensor-driven, cyberphysical manufacturing and construction. Ultimately, this opens up opportunities for novel, feedback-driven, robust and semiautonomous or autonomous construction processes that do not rely on fully preplanned models but still achieve the required performance and quality.
The long-term goal of the Cluster is to develop a co-design methodology that operates on successive levels: (a) a predictive layer that The co-design method created for this project generated the shape of each component for the pavilion according to the architectural design intent and the structural loading. 9,10 The conception and development of the transportable robotic manufacturing unit to be used was also an integral part of the co-design. 11 Only this highly integrative process made it possible to manufacture 376 different panel segments with 17 000 different finger joints in accordance with the diverse design requirements for the overall structure and its details with an accuracy of three tenths of a millimeter. Compared to solid wood segments, the hollow wood cassettes significantly reduce weight and material, but also increase the number of components by up to eight times and lead to a more complex production process. The quest for greater resource efficiency therefore had to go hand in hand with automated robotic production workflows of the shell segments.
Load-adapted and thus material-efficient shell structures are virtually extinct in today's building practice due to the high effort required for their manufacture. For the BUGA wood pavilion, a system was developed as an alternative to traditional shell construction methods in which a double-curved geometry is assembled from flat cassettes. This requires each of the pavilion's 376 cassettes to have different dimensions, which in turn is only sensibly possible with computer-aided manufacturing. For this purpose, a transportable robotic manufacturing unit was set up in a timber construction workshop. 12 One robot grips and positions the components, while the second processes them, that is, applies the glue, sets the hardwood nails for temporary fixation, mills the edges and drills the holes. On  for example, PUR foams. In addition, fiber mats are usually placed into molds with a 0/90 orientation, so that a load-adapted fiber arrangement is only possible to a limited extent which leads to a considerable amount of material waste. The construction of elaborate molds is therefore only economically viable and ecologically justifiable for series production. 13 In the construction industry however, it is usually a matter of one-offs or very small series, for which robust and simple processes are required, while high demands on precision, dead weight

| Research network multi-story building systems
This research network (Figure 8) focuses on the co-design-based development of methods, processes and systems for multi-story buildings, such as residential and office buildings, and represents the high integration challenge required in this context. Thus, the development of methods for exploratory design, optimisation and analysis, human-computer and human-robot interaction supported by augmented reality, as well as integrative data management are explored in this network in view of the particular challenges of this architectural application field, which includes divergent design drivers, multidimensional optimisation parameters, human-robot collaboration in fabrication, multi-layered data models and formats, and the associated multiple stakeholder perspectives-from the design and construction process to marketing and use. They also include the interrelation between spatial and constructive ordering systems, structural and building-physical performance as well as ecological, building-cultural and economic requirements. Methods development is directly related to the associated processes: we aim to increase the share of prefabrication processes of building elements and thereby regard structural framework and building technology integratively from the beginning.
At the same time, we enable flexible adaptation to the use-and sitespecific requirements of architecture. In this way, quality will be improved, work on the construction site will be minimized and time and cost control advanced. At the same time, control points are defined to ensure the monitoring and coordination of quality, safety and regulatory criteria. 8 This is also relevant to support seamless collaboration between skilled workers and robots in prefabrication-from the development of suitable interfaces and visualization tools to necessary expertise. 16 To this end, new methods of interaction in augmented reality 16 integrate situated visualization-that is, visualization embedded in the real environment on site. 17  The automated handling of heavy loads as well as the assembly of prefabricated elements for multi-story buildings is investigated using the example of a tower crane from the industrial partner Liebherr. 18 There are many reasons why the automation of these construction processes does not yet exist in practice. From a technical point of view, there is a lack of suitable sensor technology that meets the extreme requirements for accuracy, robustness and economic efficiency. 19 Compared to manufacturing processes in process engineer- On the other hand, assistance systems form the basis for new possible applications in assembly that go beyond simple logistical tasks. The combination of the planning of the transportation process in the digital environment and active sway damping, which ensures the desired subsequent behavior of the planned transport path, enables the independent transport of loads. In order to also take dynamic processes into account, the environment of the transported load is monitored so that the transport process can be stopped in the event of unforeseen obstacles and then replanned. 23 With the overriding goal of a higher degree of automation, new load suspension concepts are being investigated. These must be able to hold the transported load in a desired orientation despite a pendulum movement of the crane and to create additional room for maneuver in the course of positioning or assembly, for example by actively tilting the component. More flexibility in handling is to be achieved in particular by connecting further construction machines as well as cooperative load guidance. A crawler crane is used as the partner system, which operates from the ground. While the tower crane compensates for large forces in the vertical direction, the crawler crane can position components both vertically and horizontally. In combination with the newly developed load handling concept, this enables robotbased positioning in all six translational and rotational degrees of freedom. 22 Despite the desired automated assembly of prefabricated components, in the end we can only speak of a partial automation of construction. Therefore, integrative investigations into operability, adapted expertise and the best possible integration of machine and human skills in new forms of human-machine cooperation are imperative. 24 Through the use of technology, not only is human-machine cooperation possible, but human capabilities can even be increased in a targeted manner. 25 This can thus be a component for an appropriate design of workplaces of the future. The trend toward automation is happening gradually. Construction machines are already equipped with an ever-increasing number of sensors. The control hardware has also been developing in the direction of higher performance, a permanent connection to the internet for machine monitoring and tracking, as well as simple software updates over the air. These state-of-the-art favors further steps toward semi-autonomous to fully autonomous machines, which are certainly still a long way off. Complexity in this context is preferably focused on the software; in this context, many efforts have already been made in the consumer area to make it easier to use, which can certainly be transferred here. The main problem is the question of robustness, especially of sensor systems in the harsh environment of the construction industry. This criterion plays a decisive role in the selection of the sensor concept. Lastly, the economic efficiency aspect must certainly be considered, although this can easily be compensated for by efficiency increases in view of the limited costs for additional sensor technology and slightly increased hardware costs.
Construction systems for multi-story buildings are developed in direct feedback with the peculiarities of cyber-physical building processes. Special attention will be paid to the necessary flexibility and adaptability both in the initial design phase with regard to the local context of the building project and throughout the life cycle due to changing usage requirements, societal change and further technological developments. We will explore concepts of modularity, material systems and interfaces that enable new forms of spatial reconfiguration while avoiding the architectural monotony that characterizes most current approaches to serial and modular building.
In the developed world, people spend 87% of their lives in buildings, 26

| Research network long-span building systems
The second research network ( Figure 9) explores co-design-based development of methods, processes and systems for long-span buildings, which include large public, cultural, sports and infrastructure buildings. It also serves as a challenge for achieving high performance of building systems, as structural and material efficiency play a critical role in the economic feasibility, environmental sustainability and architectural articulation of long-span building systems. 27 Therefore, we aim to directly link exploratory computational design optimisation and visual analytics methods. 28 This applies both to novel, high-performance building systems and their efficient construction through end-to-end 4D modeling (spatial dimensions related to time) and the use of artificial intelligence 29 and exploratory data visualization. Visualization methods are already successfully used in many areas of industrial production, 30 but 4D modeling in construction brings special challenges. 31 One goal is to take also into account the specific characteristics of advanced manufacturing and material technology to adapt to forces at multiple scales. This integrates (a) local, graded material composition, (b) regional, force-adaptive system segmentation, connections and interfaces, (c) global shape finding and topology optimisation. flexibly adaptable actuator technology that can adapt to a wide variety of shapes of prefabricated components. Here, too, it can be seen from the development of the rapidly growing market with mini cranes and small manipulators that considerable increases in efficiency are already possible today for certain tasks, such as façade construction. This development will certainly become more widespread in the future.
The developed monitoring, sensing and machine control approaches form the basis for our investigations of semi-autonomous to fully autonomous processes for long-span construction. The use of robots plays a crucial role for these processes, in particular methods are needed for the processing of sensor data in real time, for the integration of simulation, but also for the haptic user interfaces to control the cyber-physical platform.
A critical factor for long-span structures is dead weight. In coordination with design/engineering methods and manufacturing/construction processes, we are developing novel, high-performance, lightweight, longspan building systems (C.2) based on wood-based materials and fiberreinforced composites for this purpose. The grade purity of the building materials and the monitoring of environmental impacts play an important role. A long-span building demonstrator will provide a vehicle for linking the network's research activities and evaluating its scientific results, and will also allow the associated architectural features to be studied.

| Research network building in existing contexts
The network Building in existing stock ( Figure 10 Horizontal or vertical extensions to the existing building stock place particular demands on the development of co-design-derived building systems, which must geometrically adapt to the existing boundary conditions and ensure lightweight construction to minimize the impact on existing foundations, as well as being effectively insulated acoustically and hygrothermally. The building systems should be prefabricated as far as possible to reduce noise, dirt, traffic and other impacts on the surrounding environment. A building demonstrator will also be used for this application as a checkpoint of research convergence, scientific evaluation and to assess architectural features and social acceptability.

| OUTLOOK
Digital technologies can fundamentally change the planning and building of the future. More efficient buildings and more sophisticated architectures can be expected as a result of more integrative, feedback-oriented and coordinated planning and construction processes, which at the same time make more responsible use of ecological resources. Digital approaches thus promise to solve many problems of the construction industry at the same time. They can contribute to overcoming the productivity, skilled labor and profitability crisis, but also the resource and sustainability crisis, as well as the confrontation culture emblematic for the construction sector. A new building culture that can better respond to the demands of the 21st century can be imagined through new planning methods, new ways of thinking, building materials, building processes and systems, and also through new working conditions on the building site and in prefabrication. A major challenge for this lies in the development of suitable interfaces to create universal interoperability and enable versatile monitoring and optimisation processes, so that building systems communicate with designs, robots with construction plans, construction workers with architects. This interoperability is also important to ensure that construction processes and structures remain comprehensible and flexible for all sides today and in the future, and that design processes can be corrected and calculations adjusted. In contrast, the linearity of today's approaches still imposes limits on architecture and construction technology. Mature, integrative computational processes can leave these limitations behind. In addition to the technical challenges, the nontechnical challenges and obstacles also play a decisive role. Which possibilities are taken up and further developed also depends on legal, economic and political frameworks as well as societal expectations and demands. Finally, the future of construction will also be shaped by the-especially digitally driven-concentration processes in international competition and accelerating standardization, which do not always match the new possibilities in construction technology. Also, for this reason, the considerations in the Cluster do not end with the developments in building technology, but include the requirements of the application contexts and consider the scope for action of the building professionals.