Computational Design for Multi-Optimized Geometry of Sustainable Flood-Resilient Urban Design Habitats in Indonesia

Abstract

Unfortunately, flooding is a major worldwide problem that especially affects low-lying cities like Semarang, Indonesia. Therefore, this study focuses on the flood-prone areas of Semarang, where recurring high tides and surges from severe precipitation cause havoc. In order to create water-resistant dwelling topologies, the paper explores the early incorporation of computational design approaches. Ultimately, the objective is to explore the strategic application of generative design techniques to support the development of a highly adaptive urban environment using optimization-based data-driven design approaches. With careful consideration, advanced computational methods were used to find concepts that may manage and lessen possible consequences in an efficient manner, increasing the urban landscape’s overall flexibility. Achieving the best possible solutions, which consider issues like feasibility, sustainability, durability, adaptability, and user comfort, requires the application of computational studies such as microclimatic, rainfall, energy performance, and fluid simulations. Consequently, promising advances in water retention and trajectory control features are shown by evaluations that concentrate on wind dynamics and energy considerations. One such example is GEN_8, the most optimal typology produced by additive massing approaches. In addition to showing less water retention than usual building typologies, GEN_8 optimizes energy performance to improve user experience overall. Accordingly, the computationally created geometry GEN_8’s shaded areas and facades effectively account for between 191.4 and 957 kWh/m² of yearly solar radiation. In contrast, average building typologies show higher amounts of annual solar radiation, with a minimum of 574.32 kWh/m² and a maximum of 1148.65 kWh/m². This paper’s comprehensive approach not only addresses worldwide issues but also highlights how computational design techniques may be used to construct, assess, and validate workable solutions for flood-prone locations within a flexible framework that has been painstakingly designed. As a result, the research also highlights the significance of technological advancements and computational tools in assessing, producing, and validating workable solutions for flood-prone locations by carefully curating a flexible framework that ensures efficiency, comfort, and design optimization.

1. Introduction

Regrettably, catastrophes are a global concern that surpass historical patterns and impact even areas that are typically thought to be secure [1]. These catastrophic occurrences upend communities and exceed their ability to contain the consequences of the destruction they cause. Flooding is a common and erratic natural calamity that can have disastrous effects on the environment, housing, transportation, human life, and infrastructure [2]. In fact, floods become more dangerous due to human-caused factors, especially changes in land use and the general threat of climate change [3]. Even with the implementation of certain adaptation measures, there appears to be an increasing number of disaster victims over time, particularly in locations along the Ring of Fire belt that include countries such as Vietnam, Indonesia, Japan, and Taiwan [1]. Due to tidal surges from the sea and flash floods from higher inland locations, Semarang, the capital of Indonesia’s Central Java province, stands out as being extremely vulnerable to severe flooding [4]. With its vast coasts, the larger Indonesian archipelago is home to coastal cities that are important for local commerce, housing, tourism, and fishing. In fact, the residential areas that make up 123.6 km² of Semarang pose a serious threat to the lives of the locals. Dryland farming, with an area of 68.8 km², is the second-highest land use, illustrating how supplies of food and other supplies are frequently in short supply due to a variety of factors. Moreover, the dense population in these low-lying coastal areas highlights the significant obstacles to developing and adjusting to the impending hazards of coastal flooding and sea-level rise, requiring immediate action and effective mitigation techniques [5]. In the aftermath of floods, Soetanto et al. state that it will take a long time—often years—for conditions to return to what they were prior to the flood. Additionally, numerous scholarly works underscore the critical role that community resilience plays in reducing the effects of flooding and hastening the healing process [3]. To help its citizens—especially the disadvantaged—adapt to various shocks and stressors, a resilient city—one that can bounce back from setbacks and maintain operations—is essential. Consequently, this paper advances the conversation by investigating how computational design might improve resilience in regions that are prone to flooding. Ultimately, this is achieved by introducing an atypical strategy developed into a flexible framework in which building typology serves as a trigger for the early integration of computational methodologies into design. The framework ensures a smooth multidisciplinary process that not only aids in developing a mitigation strategy against flooding but also ensures possible adaptations for sustainability measures and high user comfortability levels. Furthermore, this article attempts to build practical flood risk mitigation measures by computationally simulating a variety of house morphologies. Then, a case study is selected to test the viability, efficiency, and flexibility of the framework by analyzing a set of generated typologies. By allowing for ongoing iterations, evaluations, and modifications based on the initially identified data as well as feedback and performance assessments, the proposed framework thus seeks to accommodate various design optimization-based approaches while guaranteeing that the final design satisfies the intended objectives and criteria. The evaluation of these morphologies’ viability as affordable, water-resilient, and sustainable living spaces in Semarang, Indonesia, is made easier by this novel use of computational analyses and real-life contextual simulations [6]. Overall, this project aims to provide important insights into the creation of resilient urban environments by providing and testing flexible, efficient, and user-based computationally generated solutions in the face of rising flood hazards using this multimodal method. In a nutshell, this article supports the following:

-

Carrying out an in-depth background study on Semarang, with a focus on flood triggers, current mitigation strategies, and community life.

-

Early incorporation of computational design techniques to ensure practicality, sustainability, and occupant comfort during the design phase through the creation, analysis, filtering, and validation of several building typologies for flood prevention.

-

Evaluation of the provided design forecasts’ influence through comparison of the outcomes with traditional building typologies.

-

Stressing the value of integrating computational tools and investigating the wide range of options available with this design methodology.

2. Research Background

2.1. Triggers and Mitigation Measures

Due to urbanization and climate change, floods are becoming more frequent and intense, which presents serious worldwide hazards [7]. Due to its low-lying coastal location, Semarang is extremely susceptible to flooding, particularly during high tide events that are made worse by intense rains. Unfortunately, in such cases, rainfall increases the amount of water entering rivers and drainage systems that are already saturated [8], which overflows onto streets, fields, and residential areas. Similarly, high tides prolong and worsen the flooding in Semarang by preventing floodwaters from discharging into the sea. Acknowledging the need to combat this threat, thorough mitigation strategies and resilience-focused modifications are essential. Moreover, Figure 1 provides an overview of the different triggers that cause floods, with the three primary contributors being hydrological, meteorological, and human factors. Consequently, a series of disastrous problems are brought on by this increased frequency, including decreased crop production [9], harm to the environment and economy [10], and, regretfully, the loss of life [11]. As a pressing interdisciplinary issue, flood resilience is significantly influenced by the design decisions of architects and planners, impacting the overall building life cycle [12]. Furthermore, the complex network of different combinations of mitigation strategies and solutions can be used, particularly in the early stages of design, to reduce or manage the effects of flooding incidents. Therefore, through a holistic approach, societies and communities can enhance their preparedness for floods and mitigate their effects. Moreover, by computationally creating, evaluating, and confirming all potential outcomes, mitigation steps are made to safeguard lives, foster resilience, and ensure livelihoods. Ultimately, the goal is to achieve the most optimal solution not only in terms of resilience but also in communal and comfortability adaptation. Importantly, Warsilah draws attention to how poorly resourced and ill-prepared major Indonesian cities—like Jakarta and Semarang—are for the effects of climate change, making them vulnerable to urban sprawl and related complicated issues. To counteract the disadvantages of urban expansion, local governments—Semarang being one example—are proactively regulating land-use planning, building public mass settlements, and guaranteeing population mobility [13]. As shown in Figure 1, reducing the effects of floods typically necessitates a dual strategy that combines structural and natural-based approaches. Commonly, natural-based methods reduce the intensity of floods by using sustainable environmental techniques like soil control and afforestation. On the other hand, structural measures consist of engineered solutions that are intended to regulate and redirect flooding, like levees, dams, and management systems. A critical worldwide necessity arises in light of the increasing frequency and intensity of floods, which are exacerbated by the combined effects of urbanization and climate change. This research aims to address this issue by introducing a flexible and adaptable framework that emphasizes massing concerns in the early design phases. Consequently, this enables further modifications to tailor the proposal to the intended design profile achieved primarily, in this case, by the deliberate redirection of water flow through creative construction typologies. Thus, the principal aim of this initiative is to proactively address and mitigate the increasingly severe repercussions of these significant flood events while also accounting for occupants and their comfortability.

2.2. Computational Technologies and Selective-Based Design

Within the ever-evolving field of technology and design, the combination of selective-based design methods and computational technologies has become a game-changing catalyst that affects how engineers and designers think, develop, and optimize solutions. Research shows that 20% of design decisions are critical in the early phases and have a major influence on the remaining 80% of design decisions [12]. Hence, optimizing designs with flexibility to accommodate different occupant needs is made easier by utilizing computational tools, especially in the early stages of design. With the ability to produce numerous design iterations, the optimal solution is certain to be found through rigorous testing, evaluation, and verification against carefully chosen criteria and regulations, rather than settling for partially optimized solutions [14]. In a notable study, Maksoud et al. developed an Islamic parametric elevation to address issues with sun radiation, thermal comfort, and visual comfort. Overall, the main goal was to create an elevation with better environmental performance. Therefore, shadow studies, wind diagrams, and sun radiation diagrams were just a few of the many factors that were thoroughly examined to determine the ideal elevation. Moreover, data on solar radiation, precipitation, sun path, energy performance, and building orientation were all considered [15]. As a result, the use of computational design techniques helps to anticipate and foresee problems before they arise. Additionally, computational design techniques were essential to improving the overall performance and experience in a particular space, according to a case study conducted at the University of Sharjah. Before actual implementation, virtual reality technology was used to anticipate difficulties and shortcomings [16]. Another noteworthy case study highlighted the need to investigate wind patterns in urban areas when performing spatial interventions. Carried out in the Novi Bezigrad district of Ljubljana, Slovenia, under the direction of Lavtižar, the study evaluated many building typologies using model simulations in a virtual wind tunnel to determine which had the best ventilation capacity for future neighborhood improvements, as shown in Figure 2 [17]. When analyzed, Figure 2 in the simulation process represents a careful fusion of several parameters such as object type, size, aspect ratio, aperture width, and number of openings. Moreover, iterative adjustments and computational control are applied to these factors until the desired results are obtained. Consequently, the trustworthiness of the results is increased by this iterative technique, which guarantees the correct representation of the intricate relationships among variables. Eventually, a thorough grasp of the system is made possible by the ongoing parameter optimization, which promotes wise decision-making. Upon examined cases, research findings, and filed contributions, using computational design techniques to create a resilient city does not require an intricate design or a strict solution with set parameters, despite popular belief. Instead, by meeting the required standards and requirements, computational design promotes diversity and uniqueness in the design process enabling the person to conduct assessments and execute changes smoothly, accurately, and purposefully. In short, computational design is an instrument that allows engineers and designers to generate, analyze, and model designs in the context of actual circumstances, allowing for the comparison of outcomes across multiple scenarios. Early urban planning stages for flood-prone areas benefit greatly from the integration of computational design techniques since it takes several important factors into account. Firstly, the incorporation of computational design helps forecast the size, depth, and potential behavior of floods by enabling accurate modeling and simulation of flood dynamics. Significantly, this integration makes it possible to pinpoint high-risk areas and create specialized flood mitigation plans. Similarly, computational design aids in the development of optimal solutions by taking into account several variables including affordability, environmental effect, and community demands. Ultimately, the second method of integration is specifically highlighted in this research, with a focus on how computational design methodologies might improve urban landscapes’ sustainability and resilience to flooding. Overall, computational design thus guarantees that designs are optimized within the limitations related to each project [18].

2.3. Applications

Several case studies were analyzed and assessed to gain a deeper understanding of the ways in which different factors—such as climate, location, severity of the situation, materials availability, cultural background, and practicality—influenced the approaches that were considered to be taken in order to resolve each issue. In their comprehensive study, Ke et al. utilized a problem-centered strategy to direct a cyclic pursuit of objectives in their comprehensive study, which utilized the Design Science Research Process (DSRP) methodology. Through analysis and reflection, the data were placed within the framework of current theory in an iterative process designed to produce recommendations. In order to create a design framework, the article employed an engineering design approach and concentrated on flood resilience in urban and architectural design. This iterative process involved evaluating and reflecting on results within existing theory to generate suggestions. Generally, the study focused on the campus of Jiangxi Arts and Ceramics Technology Institute in China, analyzing the architectural arrangement, planning pattern, and details of the building for vulnerabilities. The campus, located in Jingdezhen City, Jiangxi Province, is challenged by the subtropical monsoon climate, marked by heavy precipitation, erratic rainstorms, and the possibility of geological collapse. Moreover, the vulnerability analysis examined building performance in simulated flood and rain situations, as well as the complexity of the terrain, as shown in

Table 1. Design phase simulations using Grasshopper.

Phase	Iteration Number 20 (I = 20)	Iteration Number 40 (I = 40)	Iteration Number 60 (I = 60)
Sketch
Preliminary Design
Design Reinforcement

. Specifically, the buildings to the south and northwest showed deficiencies that could have been caused by the insufficient vertical typology that impeded natural water flow. Frequently, such deficiencies happen as a result of architects neglecting to take the topography into account at the sketch stage, instead concentrating solely on the building’s function. By modifying the site elevation and adding green and gray infrastructure, Ke et al. optimized the building shape during a design refinement step. Despite certain structures remaining susceptible, the studies showed improved performance overall. In flood-prone areas, the study urges minor alterations to reduce the risk of flooding and highlights the importance of taking terrain into account when designing buildings [12]. Through the application of the learnings from these cases, the research seeks to improve the understanding and application of the analyzed strategies and scenarios in a broader context, promoting significant contributions to the field.

3. Materials and Methods

When a natural disaster strikes, the local population must decide whether to escape the damaged area, stay and participate in disaster mitigation efforts, or stay in the affected zone without taking any action [19]. Although there are many reasons why people choose to stay, two main ones that stand out are strong familial ties and financial necessity. Whilst this phenomenon is important, there are very few empirical studies that focus on communities that decide to stay in climate change hazard-prone areas, which makes these groups an interesting research topic [20]. The scientific discussion of the problems caused by flooding and land subsidence has yielded significant findings, as

Table 2. Overview of relevant research contributions on Semarang’s flooding and land subsidence.

Researcher	Year	Methodology/Approach	Result(s)	Reference
Hakim et. al.	2023	Utilization of Improved Combined Scatters Interferometry with Optimized Point Scatters (ICOPS) and computational analysis	Time-series InSAR deformation map, Land subsidence susceptibility map.	[21]
Irawan et al.	2021	Raster-based inundation model (LISFLOOD-FP)	Flood hazard map, extent and depth of flood map	[22]
Husnayaen et al.	2018	Multi-sensor satellite data, including Advanced Land Observing Satellite (ALOS), Phased Array type L-band SAR (PALSAR), Landsat TM, IKONOS, and TOPEX/Poseidon.	Vulnerability rank of the shoreline, coastal slope, geomorphology, significant wave height, mean tidal range, and sea level rise, as well as land subsidence distribution map	[23]
Nugraha et al.	2015	Vulnerability Capacity Analysis (VCA)	Tidal floods hazard map, tidal flood vulnerability map, tidal flood capacity map, tidal floods risk mapping,	[24]

illustrates, explaining a variety of strategies and solutions for these pressing problems. Therefore, this emphasizes the need for more investigation and study of the dynamics of human decision-making in early design phases during natural disasters, especially considering the risks associated with climate change as well as possible outcomes of the design within the context.

To effectively address the complex difficulties of flood resilience and urban design, a comprehensive approach is incorporated into the research methodology used in the study of computational design for sustainable flood-resilient urban habitats in Semarang, Indonesia. Fundamentally, the approach places emphasis on the creation of affordable, durable, and flexible habitats that can handle environmental issues, cater to human requirements, and tackle the consequences of climate change, particularly with reference to flooding. Furthermore, this is accomplished by applying computational design approaches heavily in the designing process, which makes it easier to conceptualize and produce design solutions that are resilient to flooding. A key component of this method is iterative testing and optimization, which enables the development of solutions that can withstand and reduce the dangers associated with flooding. Moreover, the technique heavily emphasizes data-driven design approaches, which use data to enhance solutions for sustainability, user comfort, and resilience. The adaptability and coping mechanisms of suggested building masses and typologies are thoroughly assessed in response to actual disaster scenarios using simulations and assessments. Thus, this methodology heavily relies on interdisciplinary collaboration to generate creative solutions that incorporate a range of viewpoints and areas of expertise. Moreover, Figure 3 clarifies the fundamental idea of iterative advancement inside the framework. After the identification phase begins, the process proceeds into the ideation and conceptualization stage, which is mostly dependent on meticulous simulations and analysis. These ongoing iterations, which emphasize data-driven building morphology, performance, environmental considerations, and user-centric innovations among other variables, are continuously informed by these inputs and tailored to correspond with the study objectives [25]. After parameter testing and validation, the design is updated and refined further, or iteratively repeated, until the most optimal solution is reached.

While the obtained building typologies are considered initial and generic mass prototypes of functional buildings to some extent, the primary goal, aligning with the research objectives, is for the proposed mass to serve as a functional catalyst for the surrounding urban environment, enhance energy performance, and improve user comfort. To improve the design process and carry out environmental microclimatic assessments for the application of the framework through a chosen case study, a range of plug-ins for Rhinoceros 7 through Grasshopper, such as the Ladybug Tools, were included. First, several environmental microclimatic assessments were carried out with the use of Ladybug Tools, a collection of plug-ins that make environmental design easier. The Ladybug tool typically includes ladybug, honeybee, dragonfly, and butterfly, each of which is utilized to produce a variety of environmental-related findings. Therefore, weather information, including temperature, wind speed, humidity, and radiation, was collected by the Ladybug plug-in and gave important new information for the first evaluation and comprehension of the site. Then, the optimized-based design concepts were created using Evomass, an integrated tool for performance-based building massing design, in compliance with Semarang requirements and norms. In order to guarantee flood mitigation while taking ventilation and occupant comfort into consideration, these proposed generations underwent extensive testing utilizing Computational Fluid Dynamics (CFD), finite element analysis, and thermal simulations. Hence, the optimal building typology was identified through analyses and simulations, and the results were validated by comparing computationally generated typologies with a typical Semarang housing typology. Figure 4 presents a thorough strategy to address flood-related concerns by demonstrating the procedures and overall workflow of the research.

3.1. Study Area

In flood-prone regions like Indonesia, where communities adapt to annual flooding patterns [26], cities such as Jakarta, Surabaya, Bandung, Banjarmasin, and Semarang grapple with heightened vulnerability exacerbated by inadequate resources [1]. For instance, Kelley and Prabowo’s study focused on challenges in a flood-prone village in Southeast Sulawesi, revealing difficulties accessing water, food, and damage to crops, property, and infrastructure [26]. As illustrated in Figure 5, Semarang’s coastal location makes it vulnerable to tidal flooding, erosion, land subsidence, and increasing sea levels. Generally, calculating risk levels often entails using a methodical approach to determining the possibility and possible consequences of different risks. Hence, risks related to a specific location are first assessed in various categories, including financial, operational, legal, or environmental aspects, as in this instance. Then, the possibility of each risk materializing and its possible consequences are assessed. Finally, impact and likelihood are frequently rated as low, medium, or high. In fact, the likelihood and impact scores from these assessments are then multiplied to obtain the risk level for each issue that has been identified. Risks can be prioritized according to their levels, which makes it possible to focus resources and attention on the most important risks. As a result, local and governmental activities are launched to address these challenges. Expanding upon previous endeavors, the study suggests utilizing computational design techniques to generate a variety of building types for flood mitigation, taking into account the particular difficulties and continuous projects in Semarang. Beyond just identifying the problems with flooding that exist today, the research investigates Indonesia’s architectural history to find innovative and long-lasting solutions for metropolitan areas that are vulnerable to floods. In fact, the article takes its cues from traditional Indonesian architecture, which historically prioritized functionality over style because of resource limitations. Thus, this paper aims to contribute to sustainable urban development by integrating innovation and legacy, providing practical solutions firmly anchored in the nation’s architectural and cultural identity as well as adhering to common material usage, among other significant factors.

By combining short- and long-term operational and adaptive methods to address present issues and foresee potential risks, a holistic approach is needed to build a resilient metropolis that can endure severe flooding episodes. One of Java’s biggest and most culturally diverse cities, Semarang, is characterized by architectural marvels that reflect its rich culture and heritage. Moreover, Semarang also enjoys a strategic location near major international hubs like Tanjung Mas Port and Ahmad Yani Airport. Although the city has a lively atmosphere, it faces significant hazards, particularly from tidal flooding, a persistent hazard that presents unresolved issues. This article concentrates on the second-most flood-prone area in the city, the Tambakharjo district in west Semarang, as seen in Figure 6. By utilizing an efficient design methodological framework, the study attempts to produce and investigate adaptable design alternatives for typologies as simple masses that provide additional improvement and experimentation as per the case, with the ultimate goal of reducing floods and harm to inhabitants while considering factors such as energy performance and user comfortability.

3.2. Site Evaluation

In the pursuit of comprehending the environmental context and climate nuances of Semarang, Indonesia, this study delves into microclimatic assessments. The integration of several tools facilitated the evaluation, generation, and visualization of a range of outputs by utilizing the Grasshopper plug-in. Initially, microclimatic assessments were conducted with the use of ladybug tools, especially the ladybug plug-in. Given Semarang’s well-known hot, humid, and cloudy climate, the study employed a comprehensive approach that enabled the successful completion of an efficient building performance assessment. A detailed examination, including the presentation of an annual wind rose chart and the script used, shown in Figure 7, showcased the parameters and prevailing wind patterns in Semarang, respectively. Notable results showed that the average wind speed is 3.025 m/s, mostly from the northwest as shown in Figure 7a, and that the wind is rather calm. Interesting variances were noted, including a 2.44 m/s south-easterly breeze, despite an annual average wind speed of 5.95 m/s. Although it is an extremely low possibility of occurring, a maximum wind speed of 11.80 m/s was recorded from the west-northwest and north-northwest directions. Furthermore, as shown in Figure 7b, the script interconnected Ladybug components with EnergyPlus meteorological data, enabling a detailed analysis of wind roses and offering important insights into the city’s distinct climate.

Moreover, the study carried out a thorough examination of solar radiation with the goal of estimating how much solar energy, accounting for skylight conditions, reaches the location. With that, three separate categories of solar radiation are included in this analysis: global solar radiation, diffused solar radiation, and direct solar radiation. While diffused solar radiation accounts for the solar radiation that reaches the Earth’s surface, considering it is subjected to several obstructions, direct solar energy reaches the surface directly. Hence, global solar radiation is the result of combining these two forms, giving a more accurate picture of sun exposure. Figure 8 in the analysis showcases global solar radiation using a rose chart and heatmap along with the script utilized. As can be seen from the accompanying heatmap shown in Figure 8a, the months of August through early November have the highest amounts of solar radiation, especially around midday. In contrast, late December to early January reflects the lowest solar radiation at an annual average of 516 kW/m². Moreover, Figure 8b rose chart illustrates Semarang’s propensity to have solar radiation over 1100 kW/m² in the northeast during the warmer months, with plenty of radiation reaching 800 kW/m² in the northwest. Additionally, Figure 8c presents the script detailing the necessary components for the ladybug code, including the location file for the weather map. Thus, this thorough study of solar radiation provides insightful information about Semarang’s solar radiation distribution and intensity, facilitating well-informed decision-making for a range of uses.

With two distinct seasons—the dry season, which runs from April to October, and the rainy season, which runs from November to March—Semarang is well-known for its two distinct climates. A comprehensive examination of the dry bulb —the ambient air temperature without taking into account the effect of air moisture— temperature throughout the year reveals noteworthy patterns. With an average maximum temperature of about 36 °C during the dry season, October is the hottest month according to the temperature heatmap shown in Figure 9a. Conversely, July, with a relatively low average of 21.76 °C, emerges as one of the coldest months. With that, the lowest recorded annual temperature is 18.20 °C, albeit rarely observed, especially during early mornings. Figure 9a also demonstrates that peak temperatures occur in the afternoons and nights, while mornings register the lowest temperatures. When one considers the impact of variables such as humidity, as shown earlier, it is evident that actual comfort levels may deviate from the data that is gathered. Consequently, it is essential to attend to occupant comfort concerns and improve energy efficiency in planned buildings. Furthermore, the script and Ladybug components used in the code are shown in Figure 9b.

Semarang, a coastal city in Indonesia, faces a significant risk of tidal floods and heavy rain-induced floods, posing serious threats to its residents. Regretfully, people believe that the city is not adequately ready for these kinds of disasters. Damage from tidal flooding can vary greatly, affecting houses and mosques to the point where they become uninhabitable. Tragic incidents have occurred in Semarang, causing great mental and physical suffering, family relocation, property loss, and the untimely deaths of many people. Understanding the long-term consequences of water levels in the area is essential to addressing these issues and preventing more damage. Therefore, a flooding map that shows the amount of inundation at water levels of 0 m, 1 m, 2 m, and 3 m is presented in Figure 10. Especially noteworthy is the map at zero meters, which shows current submersion and matches projections that Semarang, particularly around the shore, may be completely submerged by 2030. Furthermore, Figure 10 highlights the rising tendency of land submersion at different sea levels. Even though the maximum water level ever recorded is just around two meters [13], some interviews point to a worst-case depth of three meters. In such cases, buildings that are not built properly could engulf entire communities by floodwater. As a result, it is imperative to implement measures to enhance preparedness and mitigate the potential devastation caused by rising water levels in Semarang.

3.3. Experimental Approach

A thorough examination of the consequences and influencing elements of the design was carried out to produce a resilient building design that could reduce flood damage, ensure optimal energy performance, and enhance user comfort. After rigorous study and conception, the computational step of creating design solutions targeted at accomplishing this goal was initiated. The primary focus of the design is to employ computational methods in constructing a robust building, thereby minimizing the impact of floods through the building’s typology. Beyond merely building structural resilience, the ultimate purpose of the research is to improve Semarang citizens’ quality of life and communal experiences while reducing the danger of natural disasters. Thus, in addition to satisfying community demands, the suggested solutions also put comfort and social sustainability first. Over a hundred design iterations were produced using Evomass, an optimization-based design exploration tool, to generate the mentioned typologies [25]. The goal was to determine which design was best suited for the assigned location. Considering the robust mass and its capacity to accommodate the specified goals and objectives, careful attention was given to various considerations. Generally, these considerations span from structural and material usage to exploring theoretical potential enhancements that can be implemented without altering the silhouette of the typological mass generated where the aim is to ensure the mass aligns seamlessly with the cultural and overall design profile, considering a range of aspects. While social housing primarily focuses on finding low- and middle-class individuals at an inexpensive price, the homes must adhere to government regulations, which are further outlined in

Table 3. Minimum standards of a healthy house as per the government of Indonesia [28].

No.	Space Used	Minimum Width (Meters)	Minimum Area (Meters)	Ceiling/Plafond Height (Meters)
1	Main bedroom	3	9	+2.40
	Children bedroom	2	6	+2.40
2	Lounge and dining room	3	12	+2.40
	Join	2.5	7.5	+2.40
	Separated
3	Kitchen	1.5	3	+2.40
4	Bathroom
	Square	1.5	2.25	+2.00
	Lengthwise	1	2.70	+2.00
	Separated	1
	Bathroom	1	2.10	+2.00
	Water closet		1.40	+2.00

.

Thus, a thorough analysis of Semarang’s construction regulations and standards was carried out to direct the design process, defining specifics such as built-up areas, levels, and bedrooms. Likely, the development of these design recommendations was guided by a set of criteria that prioritized durability, robustness, and viability while considering regularly used materials, among other important factors. For instance, to comply with Semarang’s construction rules, one- to two-level constructions were taken into consideration while examining modifications for intermediate- to high-income structures. Furthermore, a maximum span of 6 m without cantilevers was established, assuming that wood would be used for construction. From the extensive generative process, ten carefully selected building typologies were chosen as illustrated in

Table 4. The selected generated building typologies.

Subtractive
	GEN_1	GEN_2	GEN_3	GEN_4	GEN_5
Additive
	GEN_6	GEN_7	GEN_8	GEN_9	GEN_10

, representing the most promising solutions among the initially obtained generations. These typologies serve as focal points for further investigation due to their diverse features. The provided typologies serve as a starting point for further investigation and improvement, guaranteeing that the finished design will not only satisfy structural specifications but also improve the quality of life for those living in Semarang.

In the utilization of Evomass, both the additive and subtractive form massing components play a pivotal role in shaping architectural masses. Although these components’ initial setup pages are roughly similar, as seen in Figure 11, their differing approaches to mass computation produce very different results. Additive form massing, as its name suggests, combines several masses to create elaborate and multi-layered constructions. By removing from the overall mass, the subtractive form massing component, on the other hand, generates the final mass and is frequently used to construct subtle variations of Semarang courtyard-style dwellings or communal spaces in the given situation. Each component was strategically chosen to yield a distinctive mass, emphasizing specific functions. In contrast to the conventional courtyard setting, the additive form component, for example, made it easier to create clusters with useful community spaces on terraces. Conversely, variants of Semarang courtyard-style dwellings with minor alterations were produced using the subtractive form component. The overarching goal was to develop a variety of house typologies that broke subtly from traditional rectilinear architecture. Although the produced designs might not fully utilize the tool’s massing capabilities, this restriction is deliberate given the structural and construction limitations inherent in the design morphology. Table 4 illustrates the additive form massing component’s contribution to the final five generations, while the subtractive form massing component influenced the first five generations. A general summary of the restrictions, spans, and other factors considered for each technique used to create the architectural type is in Figure 11.

3.4. Analyses and Simulations

The deliberate selection of ten distinct building typologies was made with great care and thought, adhering to regulations and standards to guarantee a thorough portrayal of all available architectural alternatives. Overall, the goal was to determine the most resilient and effective flood mitigation typology that simultaneously reflects optimal energy performance and achieves user comfort, considering various noteworthy aspects. After identifying the ideal typology, it can be further refined to emphasize aspects of the Semarang-based Indonesian way of life, like material consumption and social sustainability, without substantially changing the mass that was first acquired. Through a series of simulations and assessments, the adaptability and coping mechanisms of the proposed building masses and typologies in the face of real-life disasters encountered in Semarang were comprehensively evaluated. Therefore, this thorough analysis seeks to ascertain the projected generations’ viability while considering their capacity to endure and adjust to difficult situations. By pursuing a holistic approach, it was ensured that the typologies selected would not only mitigate the effects of flooding but also blend in smoothly with the distinctive features of the local way of life and stay robust when faced with unforeseen obstacles.

Generally, design choices for flood management are often informed by wind assessments and energy performance models, among other considerations. Through the optimization of several elements such as building orientation, ventilation systems, and energy efficiency measures, these assessments improve the overall resistance of structures to extreme weather events like flooding. Therefore, architects can strategically place buildings and incorporate measures that reduce flooding risk while enhancing occupant comfort and safety by having a thorough grasp of wind patterns and energy dynamics. Buildings’ structural resilience is bolstered by this comprehensive strategy, which also advances the more general objective of developing resilient and sustainable urban settings that can survive the difficulties brought on by climate change.

3.4.1. Water Retention Analysis

As the main hazard to Semarang, the analysis gave priority to assessing floods caused by rainfall. In order to comprehend how building façades or outskirts affect rainfall direction, the investigation considered building typologies in their simple conceptual massing phase as the initial stage. Significant insights into water flow, trajectory, and retention were found when ten chosen proposals were tested at different water flow intensities. As indicated in

Table 5. Water-flow analysis for the different building typologies at various intensities.

	Low	Medium	High
GEN_1
GEN_2
GEN_3
GEN_4
GEN_5
GEN_6
GEN_7
GEN_8
GEN_9
GEN_10

, which further reflects the range of predicted rain intensities in Semarang, the ten generated typologies that were selected for testing were put to the test at three different intervals of water flow intensity. Every generation was assessed to gain a deeper understanding of the role that building shape has in regulating water flow, the stream’s trajectory downstream, and water retention, among other aspects. As a result, the investigation showed different water flow magnitudes along red and blue lines, excluding GEN_1 to GEN_4 because of their centric-courtyard design and significant water retention capacity. Overall, this consideration is grounded in the standards and requirements derived from pre-identified land plots, as well as the minimum requirements for dwellings in Indonesia, further indicating that the ground floor is most efficient when occupied. Moreover, the study acknowledged the possibility of variations in results with different building orientations in upcoming assessments and simulations, even though the stream patterns in individual studies were consistent.

3.4.2. Waterflow Analysis

The subsequent stage investigated the contextual behaviors of selected geometries, distinguishing between individual dwellings and townhouses. Conducting tests with two orientations highlighted the significance of typology in rainfall redirection, while also recognizing the impact of varying alignments. Reflecting on the first analysis findings, the six selected geometries were tested again for typological impact on stream course, as shown in

Table A1. Water flow analysis for the different building typologies in a contextual reference.

	Separate Houses		Townhouse
	Orientation_1	Orientation_2	Orientation_1	Orientation_2
GEN_5
GEN_6
GEN_7
GEN_8
GEN_9
GEN_10

, omitting those that were previously found to be prone to water retention. As noted, water collection was shown in GEN_5, GEN_6, GEN_7, and GEN_10, highlighting the significance of geometry in forming narrow nocks that affect water behavior. Moreover, water redirection with limited collection was significantly highlighted by GEN_8 and GEN_9, which is inferred to have been impacted by overall typology. Furthermore, the study acknowledged that various alignments and orientations may have an impact and that a more comprehensive understanding of typological behavior was necessary. For example, GEN_9 demonstrated how changing geometric orientation could have a substantial impact on water collection, highlighting the significance of thorough studies in establishing a building’s resilience to floods. In contrast, it is evident that certain adaptations and layouts of GEN_5, for instance, do not achieve the same level of efficiency observed in the results of other variations.

Generally, the script used to carry out the water-flow analysis on the created building typologies is shown in Figure A1. The analysis used a charged point to control the flow dynamics while simulating the water stream using a swarm with a field technique. With that being said, making a mesh out of the closed brep of the geometry was the first step. Then, plug-ins such as Anemone and Weaverbird made the analysis’s execution and calculation easier. Finally, the water stream’s curves were color-coded to precisely represent the different flow magnitudes to improve visual representation. Ultimately, this method ensured a comprehensive and visually informative assessment of how water would interact with the proposed building typologies.

3.4.3. Shadow Analysis

As mentioned, this research stresses optimizing energy performance for sustainability, comfort, and feasibility in addition to developing flood-resistant buildings. Therefore, solar hour analysis, which evaluates the building’s solar energy utilization efficiency, is one of the main components analyzed at this stage. Although elements such as overhangs, doors, and apertures may affect the outcome of the analysis, this study particularly looks at the building’s silhouette to see if any shade is projected through the mass. In essence, the research aims to assess and emphasize the viability of the general typology in flood mitigation, considering various environmental factors that can influence both energy performance optimization and users’ comfort perception. Thus, the methodological approach is developed within a flexible framework, allowing for future enhancements and modifications to the generated proposal while maintaining the overall silhouette of the mass due to the rooted functional efficiency of the mass established during earlier design phases. Generally, this approach precedes architectural decisions and design strategies, enabling a thorough analysis phase customized to criteria predetermined based on challenges and opportunities identified through a flexible framework. Moreover, the main objective is to gain a thorough understanding of the interactions between the environment and the building type in order to optimize the overall mass for increased sustainability, comfort, and practicality. To determine if it would be feasible to create outdoor recreational areas and community services that would complement the cultural and lifestyle preferences of Semarang’s citizens, the earlier two building types were carefully considered. Consequently, shadow analysis was conducted to understand the impact of different building typologies on sunlight exposure as tabulated below in

Table A2. Shadow analysis.

	Single Dwellings		Townhouse
GEN_5
GEN_6
GEN_7
GEN_8
GEN_9
GEN_10

. Independent houses, especially in GEN_6, GEN_8, and GEN_9, dramatically reduced direct solar exposure, according to the data from the six planned generations. Conversely, townhouse layouts like those in GEN_7 and GEN_10 created more private spaces that were appropriate for community functions by forming shaded pocket spaces. Because of this, townhouse designs in Semarang can be modified to provide areas that are shaded and less exposed to the sun, which makes them ideal for outdoor recreation and community services. The results, which are displayed in Table A2, show that different generations have varying maximum ranges of direct sun exposure for different home adaptations. Notably, GEN_6, which has a separate housing adaptation and a smaller maximum range, records an annual maximum average of 4213 h of direct sun radiation. By contrast, GEN_9 has a greater maximum average of 4295 h annually with independent housing adaptation. However, GEN_10 indicates a lower maximum average of 4176 h yearly when townhouse building modification is considered. Meanwhile, for single dwellings, GEN_8 is ranked the second typology variation with a lower maximum average of 2349 h annually. In contrast, GEN_9 under the townhouse adaptation scenario gets the highest yearly maximum average of 4264 h. While acknowledging the potential for further enhancements in each typology, the primary focus of the research at this phase is to identify the most effective building typology based on initial massing, with implications for practicality among other considerations. All in all,

Table 6. Ground direct solar hours data summary.

GEN	Single House	Townhouse
GEN_5	0–4285	0–4208
GEN_6	0–4213	0–4201
GEN_7	0–4283	0–4193
GEN_8	0–4270	0–4208
GEN_9	0–4295	0–4264
GEN_10	0–4283	0–4176

Lower Maximum Range. Higher Maximum Range.

offers a thorough overview of the results.

3.4.4. Direct Solar Radiation

An examination that goes beyond shadows is necessary since the evaluation of the shadows created by suggested architectural typologies is recognized as an inadequate metric by itself. Further research was carried out to improve comprehension of the building typologies under consideration by examining direct solar exposure and solar radiation. This involved assessing the advantages that each typology provides in preserving energy-optimized performance and determining elements that can be combined depending on the initial interaction with the environment. As can be seen in

Table A3. Direct solar hour on the generated proposed building typologies.

	Single Dwelling	Townhouse
		Left	Middle	Right
GEN_5
GEN_6
GEN_7
GEN_8
GEN_9
GEN_10

, the results of the analysis of the direct solar hour reveal different levels of solar exposure. Top views, which correspond to the roof in most cases, generally exhibit the highest levels of exposure, except for GEN_5, which has a recommended balcony as indicated. Furthermore, the script presented in Figure A2 outlines how computational evaluations considered the impact of surrounding buildings to provide context for the scenes that were examined. As seen, most elevations receive an average of 1988 h of sunlight per year, despite a maximum exposure of 4418 h per year. Moreover, key observations and modifications seen in the plans under examination are summarized in

Table 7. Direct sun hour data summary for the elevations.

GEN	Single House	Townhouse (Middle)	Townhouse (Left)	Townhouse (Right)
GEN_5	1104.5–2429.9	0–2650.8	0–2650.8	0–2650.8
GEN_6	1104.5–2650.8	0–2650.8	0–2650.8	0–2650.8
GEN_7	883.6–2650.8	0–2650.8	0–2650.8	0–2650.8
GEN_8	662.7–2429.9	0–2650.8	0–2650.8	0–2650.8
GEN_9	662.7–2650.8	0–2650.8	0–2650.8	0–2650.8
GEN_10	883.6–2650.8	0–2650.8	0–2650.8	0–2650.8

Lower Maximum Range. Higher Maximum Range.

. Without a doubt, sun exposure understanding is essential to comprehend how shadows and light are distributed throughout a building and forecasting possible glare or heat concentrations in interior spaces. Interestingly, masses produced by the subtractive mass form components always result in the most shaded patches on the elevation, as seen in GEN_5. In spite of this, the selection criteria used in previous stages showed more prominent results so far in favor of typologies generated through the additive form massing components, leading to generations with significant proportions of shaded areas, elevations, and corners of junctions between masses. This additive method, which complies with the selected criteria and advances our understanding of shade dynamics in the examined building typologies, makes use of extruded masses to form pockets for possible outdoor areas, enabling a higher possibility of future iteration flexibility.

3.4.5. Solar Radiation Analysis

In order to evaluate the effects of solar radiation on the six generated prototypes that were chosen, more assessments were carried out with the goal of modifying water channels, reducing sun exposure on elevations, for potential Semarang public areas, and understanding possible enhancement that can also be further reflected on the users’ comfort. Ultimately, the chosen typology should indicate the ability of solar-related technologies incorporation for prospective energy generation in addition to demonstrating adaptability for sustainable practices. The outcomes, detailed in

Table A4. Solar radiation analysis.

	Single Dwelling	Townhouse
		Left	Middle	Right
GEN_5
GEN_6
GEN_7
GEN_8
GEN_9
GEN_10

, from a subsequent solar radiation analysis, considered varying building conditions, underscore the relevance of such investigations. As seen, solar radiation analysis plays a crucial role in energy efficiency optimization by providing guidance for building typologies about window placement and possible sustainable practices. Ultimately, this process fine-tunes the creation of building typologies, pinpointing those most likely to provide an optimal degree of energy efficiency consistent with Semarang’s lifestyle requirements.

Thus, the analysis of solar radiation in Table A4 reveals a notable extreme value of 1914.41 kWh/m² annually, primarily concentrated on the building’s top view, presumed to be only the roof, across all proposed generations except GEN_5. Significantly, GEN_5 shows different quantities of solar radiation on a balcony that the script originally created through the subtracting form component. Moreover, the incorporation of site context as a parameter for the code execution unveils that elevations receive less solar energy than initially anticipated. Logically, solar radiation levels vary depending on the mass typology; the average solar radiation level at the six selected elevations is predicted to be 670.04 kWh/m². Notably, the lowest annual maximum average of solar radiation for independent dwellings is recorded at 382.88 kWh/m² for GEN_7, GEN_8, and GEN_10, indicating a significant impact on the performance of these typologies. Also, in the adaptation of the townhouse typology, particularly for GEN_7 and GEN_8, the proposals demonstrated optimal efficiency, featuring an annual low maximum average solar radiation of 670 kWh/m² primarily on elevations while maintaining higher levels of exposure on the roof. The script, as illustrated in Figure A3, includes all of the essential components that specify the study’s location, duration, geometry, and surrounding context. Furthermore,

Table 8. Solar radiation data summary for the generated elevations.

GEN	Single House	Townhouse (Middle)	Townhouse (Left)	Townhouse (Right)
GEN_5	478.6–957.2	0–957.2	0–957.2	0–957.2
GEN_6	565.28–957.20	0–861.48	0–957.2	0–957.2
GEN_7	382.88–957.2	0–670.04	0–957.2	0–957.2
GEN_8	382.88–957.2	0–670.04	0–957.2	0–957.2
GEN_9	478.6–957.2	0–765.76	0–957.2	0–957.2
GEN_10	382.88–957.2	0–765.76	0–957.2	0–957.2

Lower Maximum Range. Higher Maximum Range.

provides a holistic overview of the collected data, providing a full viewpoint on the dynamics of solar radiation investigated in this work.

3.4.6. CFD Analysis

In pursuit of enhancing user comfort, this research employed CFD analysis on six computationally generated building proposals. Furthermore, the study sought to evaluate if turbulence might have a detrimental effect on long-term user well-being and to comprehend how different building typologies affect wind movement. Moreover, the goal of the CFD analysis, which was carried out using the GH_wind component as shown in Figure A4, was to pinpoint a geometry parameter that would hold important details about the typical wind conditions and site features. Consequently, the acquired data, presented in

Table A5. CFD analysis results.

	Single Dwellings	Townhouses
GEN_5
GEN_6
GEN_7
GEN_8
GEN_9
GEN_10

, showcased how this information facilitated the successful execution of the CFD analysis. As seen, the color-coded representation of wind speed highlighted variations among generations, with certain adaptations creating significantly calmer environments noting that blue represents minimum speed while red represents the absolute maximum and green represents the transitional speed, underscoring the significance of building typology. More specifically, the data obtained showed that single dwellings promoted wind-turbulent-free alleyways for increased user comfort. However, the townhouse adaptation showed higher wind flow rates, which might affect comfort levels if not controlled and regulated properly. When comparing the various generations depicted in Table A5, it can be seen that the different house adaptations reflect different degrees of airflow within the created alleyways. A noteworthy feature of GEN_8 that set it apart from other proposed typologies like GEN_5 or GEN_7 was its seamless integration of comfort and urban community notions. The townhouse adaptations for GEN_8 aligned closely with the pattern of wind movement, suggesting a strategic and controlled approach could enhance the humid climate’s tolerability, especially in the case of separate dwelling houses. Likewise, GEN_8 exemplified one of the adaptations that remarkably maintained a tranquil urban environment among the proposed typologies, sustaining the highest wind intervals around the building clusters. Consequently, this behavior provides a flexible template that can be tailored according to the structure’s needs or alternative functions, all while ensuring a high level of comfort in urban spaces between different typologies. Furthermore, the comparison of typologies emphasized how GEN_8, with its wind-engulfed mass, showcased the promising potential for sustainable integration and elevated user comfort.

4. Results

A wide range of criteria were used to assess how effective the computationally generated solutions are, including user comfort, cost-effectiveness, sustainability, resilience to flooding, and adaptability to changing environmental circumstances. Many factors were considered while addressing user comfort in flood-prone locations, which include preserving thermal comfort, maximizing the use of natural lighting, lowering noise levels, assuring ideal indoor air quality, and facilitating easy access to facilities and services. Overall, the goal of giving priority to these elements was to encourage the development of resilient and livable environments where people’s health comes first. Thus, after examining the computationally generated typologies against the carefully curated criteria, ten computationally generated geometries were further selected and examined in order to create, assess, and validate a building typology that improves resilience in metropolitan regions that are prone to flooding, among other factors. Moreover, the study determined that GEN_8 was the optimal choice, incorporating computational methods and simulations conducted during early-phase design decisions. Focusing on user comfort and energy efficiency, GEN_8 demonstrated promising results as a current rough typology while also demonstrating the potential for future iterations and enhancements. Therefore, a side-by-side comparison of GEN_8 with a standard Semarang typology confirms the importance of the suggested typology as shown by the data in

Table 9. Comparison between typical house typology and GEN_8—water flow analysis.

	Low	Medium	High
Typical
GEN_8

. Both typologies demonstrate efficacy in water drainage through simulations; however, GEN_8 demonstrates greater adaptability by diverting water toward the closest stream. Consequently, this emphasizes the useful advantages of the generation that is being offered, highlighting GEN_8′s capacity to build an environment that is resilient enough to endure natural calamities and support sustainable urban development.

To refine the computationally generated building typology as well as obtain data for further selection criteria refinement, extensive energy-related analyses and CFD simulations were conducted, comparing the results for the two previously discussed geometries as detailed in

Table 10. Comparison between typical house typology and GEN_8.

			Typical			GEN_8
Shadow Analysis	Single Dwellings
	Townhouses
Direct Sun Hour Analysis	Single Dwellings
	Townhouses	Left
		Middle
		Right
Solar Radiation Analysis	Single Dwellings
	Townhouses	Left
		Middle
		Right
CFD Analysis	Single Dwellings
	Townhouses
Comfort Analysis

. For instance, the analysis of shadow patterns uncovered subtle distinctions between the GEN_8 and usual typologies, with GEN_8 exhibiting a special capacity to create adaptable environments impacted by the local Semarang culture. At first glance, it might seem that the typical typology exposes the ground to a more uniform distribution of sunlight patches compared to GEN_8. However, upon considering the shadow pattern of GEN_8, the results reveal a well-rounded distribution of regions with varying degrees of shading and exposure. Ultimately, this gradual shift creates private, semi-private, and public communal areas that are adequately shaded, along with regions that are extremely exposed for utilizations such as future sustainable applications or agriculture. For example, GEN_8’s typology for townhouse structures was taken into account because of its ability to provide different elevations with strategic shade, which increases its potential for efficient energy performance [29]. Moreover, a high degree of adaptability in the areas and urban solutions that can be given and impacted by the people’s culture and way of life in Semarang is also offered by taking into account townhouse buildings constructed using the GEN_8 typology. Interestingly, the shadowed elevations of GEN_8 produced by the extruded masses were important for solar radiation studies and helped determine the locations of windows to mitigate glare and other sustainable aspects, unlike the typical Semarang building typology that offers limited enhancements in comparison. Furthermore, the CFD simulation demonstrates that the alleyway design of GEN_8 results in less wind movement, which guarantees improved user comfort in comparison to the traditional typology. During the investigation, many characteristics have come to light as critical to the development of resilient and livable environments suited to Semarang’s requirements. These characteristics include noise reduction, accessibility, natural lighting, thermal comfort, and indoor air quality. These criteria are critical for assuring comfort but also for improving residents’ general well-being and ability to withstand environmental shocks, which are common in flood-prone locations like Semarang. Through careful consideration of these variables, the research seeks to create living conditions that promote resilience, flexibility, and comfort, helping build prosperous and sustainable communities in Semarang. This innovative method stresses the possibility of rerouting and managing wind to provide a comfortable and useful urban environment in addition to taking wind circulation into account as the results indicate through the expected user comfortability values shown in Table 10.

5. Discussion

Through an extensive examination of recent research findings and a critical assessment of relevant literature, this study sought to highlight the critical role that computational design methodologies play in mitigating natural disasters at the urban scale through optimal data-driven design decisions during early design phases, with a focus on Semarang’s flooding problem as a case study. Through a holistic approach, the foundation was laid for a meticulous and flexible framework that primarily tackles aspects derived from the examination of the present challenges and requirements faced by the residents. Through this research, the key objective was to implement the aforementioned approach on the computationally generated building typologies and to gain a deeper understanding of how rough conceptual masses could be generated in accordance with predetermined sets of rules and regulations during early design phases to achieve a comparably optimized performance. Furthermore, the proposed framework facilitated a continuous sequence of generations, evaluations, and validations until the proposed design reached a level of satisfaction and optimization, surpassing countless other potential generations and current applications. This research integrates interdisciplinary perspectives from architecture, engineering, and environmental science by combining expertise in computational design, structural engineering, environmental modeling, and urban planning to develop comprehensive and innovative solutions for flood-resilient urban habitats, emphasizing the importance of cross-disciplinary collaboration in addressing complex urban challenges. Mainly, the goal of the study was the creation of a building typology that was conceived using computational techniques, with the main goal being to reduce damage while decreasing the risk of flooding. In addition to emphasizing flood resilience, the innovative chosen building design aims to maximize energy efficiency and improve occupant comfort. Thus, the overarching goal was to foster an environment conducive to communal services and urban spaces that align with the distinctive lifestyle of Semarang, as per the need of the case. During the design phase, the viability and availability of construction materials were carefully evaluated, with an initial emphasis on putting strong, durable alternatives into practice. In order to adequately reflect the changing demands of the community, a set of carefully considered and modified building rules and standards were implemented as the project progressed, such as rough considerations for mixed-use and shared communal facilities, span, rough areas, and material specifications. Ultimately, minimizing environmental impacts and displacement risks was crucial, as was prioritizing the rights and well-being of residents in the early conceptualization, design, and implementation stages. Other ethical considerations related to the application of computational design solutions in vulnerable communities include ensuring equitable access to resilient housing, protecting cultural heritage community identity, encouraging social inclusion and participation, and so on, all of which are outlined throughout the computational conceptualization process.

Utilizing advanced computational design techniques, the study generated an extensive array of variations by manipulating a predefined set of parameters, aligning with Semarang’s building regulations, codes, and standards. Crucially, the goal of this procedure was to accurately convey the cultural and historical core of Semarang while reducing the number of significant adjustments that had to be made to the originally created typology. Therefore, with the use of the Evomass tool, a Grasshopper plugin that allows for optimal mass exploration, more than 100 generations have been examined, contrasted, and eliminated according to preliminary standards that were based on preferences and expected results. Then, upon the preliminary selection phase, ten different masses were chosen from computationally generated building typologies using two different generative approaches. Using the additive form massing component or the subtractive mass form massing component, these chosen typologies included a variety of building generations, all aimed at mitigating floods by modifying watercourses, optimizing energy performance, and developing socially responsible urban public areas, among other considerations. Reflecting on the study’s main objective of integrating diverse masses to prevent flooding and improve urban living in Semarang, the interconnected set of geometries promotes community commodities and amenities. The preliminary examination, which centered on the course of water flow at different intensities to replicate actual situations in Semarang, resulted in the omission of typologies with courtyard modifications because of the issues associated with water retention. After that, a detailed analysis was conducted on the remaining six computationally created building typologies, with a focus on wind, shadow, and energy efficiency studies. Moreover, the studies examined two building adaptation scenarios: single-dwellings and townhouses, with the aim of determining the most comfortable and energy-efficient options. Out of all generations, GEN_8 stood out as being resilient and energy-efficient, reducing heat gain and glare while maintaining a high level of user comfort. Despite site limitations, various GEN_8 elevations with differing degrees of annual solar radiation permitted sustainable or solar-powered alterations. Furthermore, CFD analysis indicated well-ventilated alleyways and communal areas within GEN_8, offering comfort and adaptability through future regulation techniques. Consequently, the results of the multidisciplinary investigation are consolidated in

Table 11. Data summary for the building typologies generated.

GEN	Water Retention	Direct Sun Hour (Ground)		Direct Sun Hour (Mass)				Solar Radiation
		Single House	Townhouse	Single House	Townhouse (middle)	Townhouse (left)	Townhouse (right)	Single House	Townhouse (middle)	Townhouse (left)	Townhouse (right)
GEN_1	✔
GEN_2	✔
GEN_3	✔
GEN_4	✔
GEN_5	X	0–4285	0–4208	1104.5–2429.9	0–2650.8	0–2650.8	0–2650.8	478.6–957.2	0–957.2	0–957.2	0–957.2
GEN_6	X	0–4213	0–4201	1104.5–2650.8	0–2650.8	0–2650.8	0–2650.8	565.28–957.20	0–861.48	0–957.2	0–957.2
GEN_7	X	0–4283	0–4193	883.6–2650.8	0–2650.8	0–2650.8	0–2650.8	382.88–957.2	0–670.04	0–957.2	0–957.2
GEN_8	X	0–4270	0–4208	662.7–2429.9	0–2650.8	0–2650.8	0–2650.8	382.88–957.2	0–670.04	0–957.2	0–957.2
GEN_9	X	0–4295	0–4264	662.7–2650.8	0–2650.8	0–2650.8	0–2650.8	478.6–957.2	0–765.76	0–957.2	0–957.2
GEN_10	X	0–4283	0–4176	883.6–2650.8	0–2650.8	0–2650.8	0–2650.8	382.88–957.2	0–765.76	0–957.2	0–957.2

, which provides a thorough description of the findings.

After GEN_8 was determined to be the optimal building typology, a comparison study with a typical Semarang dwelling typology was carried out in order to validate the main premise and findings obtained. As a result, the findings showed that GEN_8 harmonizes with long-term goals, comfort levels, and cultural considerations in addition to introducing an entirely novel set of defense mechanisms. Remarkably, GEN_8’s complex layout efficiently reroutes water flow to the closest water source, demonstrating exceptional endurance in contrast to traditional building types. Furthermore, GEN_8’s many suitably shaded facades greatly improve energy efficiency, outperforming Semarang’s current traditional home typology in this regard as well. Therefore, this energy efficiency aspect is particularly crucial in addressing broader environmental concerns and underscores the superiority of GEN-8 in fostering sustainability and resilience in the face of natural occurrences.

Eventually, the article explores the application of computational design techniques in creating structures and generating building typologies for flood protection. To meet the research objectives, a meticulously created flexible and adaptable framework was developed to facilitate the integration of computational design with early design processes, aiming to foster a data-driven design approach. Tailoring this framework for application in this paper as a mitigation approach principally involved employing generated building typologies to control water trajectory. Ultimately, this adaptation retained a focus on considerations such as high-energy performance optimization and user comfortability. Despite encountering limitations that impacted the results, such as data and resource availability and integration with existing systems, the study suggests future considerations, to enhance the findings. Furthermore, solutions for water diversion and purification were suggested as ways to address Semarang’s water scarcity problems. Therefore, several other considerations can yield from conducting more research into non-structural mitigating strategies and infrastructure general enhancements. Likewise, the exploration and improvement of the typological mass involved incorporating architectural elements and implementing styles, followed by comprehensive evaluations of the performance as a result. The study underlined how crucial it is to comprehend how flexible building typologies may be within legal constraints. Hence, additional research can be performed to test out ways to seamlessly integrate shared facilities and indoor and outdoor areas. Using virtual reality technology is also recommended in future research to evaluate and improve the study’s results, offering a thorough and immersive assessment.

With that being said, the findings of the research and methods have extensive ramifications for dealing with flood-related problems, not just in Semarang but also in other cities throughout the world that are dealing with comparable problems. However, careful consideration of several crucial aspects is necessary for the successful implementation and scalability of these solutions. To make sure that suggested modifications are in line with current urban structures and processes, infrastructure compatibility is crucial. Likewise, the cooperation and support of different community members, governmental organizations, and businesses engaged in the planning and execution process, and stakeholder engagement is essential [30]. Generally, the effective implementation of flood mitigation projects necessitates enough financial resources, which calls for careful resource allocation and budgeting. The significance of continuous monitoring, maintenance, and adaptation as urban settings change is emphasized by long-term maintenance planning, which is essential to guaranteeing the continued functionality and efficacy of applied interventions throughout time.

Furthermore, there is a great deal of promise for improving urban resilience on a larger scale due to the transferability of computational design methodology, generative design techniques, and optimization-based approaches. Cities with comparable flood risks might gain from scalable and flexible solutions informed by the insights and innovations of others by exchanging knowledge and best practices. The transformative potential of computational design in building flood-resilient habitats is demonstrated in this research, which makes a substantial contribution to the discourse on sustainable urban development. The study emphasizes how crucial interdisciplinary cooperation, innovation, and community involvement are. In order to meet these demands, it presents a state-of-the-art, adaptable computational framework that provides distinct identification through specified parameters and contextual factors. As a result, this new framework sheds light on data-centric design and produces a generative solution optimized and precisely tailored to the user’s needs and preferences. Through preliminary studies of flood-prone areas, engineers and architects can easily include parameters and data into the framework, producing an almost infinite number of variations. Through analyses, simulations, and evaluations of the resulting geometries, these iterations are subjected to a rigorous examination that guarantees the pursuit of the most refined design rather than just a partially optimized solution. As such, the study emphasizes how crucial it is to incorporate computational design techniques into frameworks for urban planning. Thus, this integration improves disaster preparedness and response plans in communities that are vulnerable to disasters while supporting the development of resilient, sustainable infrastructure. For urban planners and politicians managing the complex difficulties stemming from growing urbanization and the intensifying impacts of climate change, such insights are vital.

6. Conclusions

This study delves into the significant challenges faced by Semarang, a city in Central Java, Indonesia, due to its susceptibility to frequent and severe flooding, examining current coping mechanisms and proposing innovative solutions using computational design. Drawing from contributions in the same field and an ongoing assessment of triggers imposed on Semarang, the project aims to establish a building morphology technique to produce resilient cities capable of enduring shocks and effectively managing disasters. Ten workable recommendations were carefully selected from more than 100 generations, with each one focusing on certain resources or tactics. Furthermore, to evaluate the potential of each alternative in a real-world setting, extensive simulations including CFD analysis, shadow analysis, solar radiation analysis, water flow analysis, and direct sun hour analysis were carried out. Moreover, water flow analysis-driven elimination led to the discovery of six viable solutions that not only reduced flooding but also showed high energy efficiency and occupant comfort. Following the initial phase, a subsequent series of analyses was undertaken with the primary objective of pinpointing areas with potential shading, surfaces exposed to ample direct sunlight, and potential energy-related optimizations and sustainable practices. Ultimately, the main objective was to aid in the creation of a building mass that is practical, resilient, durable, and optimized for energy performance. Consequently, this thorough analysis aimed to establish the foundation for further improvements, guaranteeing that the final structure complies with sustainable standards and successfully tackles the environmental issues presented by Semarang. With its innovative concepts and ease of adaptation for future sustainable measures, the flexible typology known as GEN_8 proved to be a noteworthy proposal compared to a typical building typology or Semarang. Most importantly, GEN_8 demonstrated significant results in redirecting water flow and exhibited a notable lack of capacity for water retention, positioning it as an intriguing and effective solution. In particular, GEN_8 brought shadow casting, which improved user experience and opened the door for the smooth incorporation of sustainable practices. Overall, this thorough study offers useful insights into designing resilient urban environments, including building masses that are feasible and well-suited to Semarang’s environmental difficulties. Therefore, in summary, this study emphasizes the following contributions:

-

Examine the importance of using computational design techniques in the early stages of the design process to create architectural typologies that are resilient, durable, feasible, and sustainable.

-

Investigate the possibility of using additive and subtractive methods to computationally generate masses to form areas, shade regions, and represent the local way of life in Semarang.

-

Conduct studies and simulations in real-world settings to make sure typologies are able to adapt to flooding events while meeting high user comfort and energy efficiency standards.

-

Use computer simulations and analysis to verify and contrast the suggested generation with typical building typologies in Semarang.

-

Assure that the framework for the generative process is still flexible and adaptive by modifying parameters in accordance with prescribed guidelines and standards.