A survey of machine learning techniques in structural and multidisciplinary optimization

Field Information Modeling (FIM)®: Beyond point cloud processing - The case of generative decision support for skeletal spatial structure systems

“…Automated point cloud processing involves the automatic assignment of the acquired points to their corresponding real-world elements as de ned in the n-D designed BIM, and/or technical speci cations. Once the points are correctly assigned, they can be used to report progress 28 , detect dimensional incompatibilities 29 , update the design BIM 28 , generate digital twins 30 , or perform generative design optimization 13 . FIM® for point cloud processing is advantageous since it incorporates all aspects of sensor characteristics (e.g., precision and calibration), construction errors, and a-priori baseline planned information into a single generic framework.…”

“…This new design may be utilized for new construction projects or to support sustainable reconstruction, renovation, and rehabilitation efforts of the existing structure when required. By virtue of its nature, repurposing existing designs as baseline to generate optimal designs in accordance with local requirements, exhibits several important advantages, some of which are: reduction of time and cost of engineering design (particularly schematic design), which constitutes between 7.9-19% of total cost for new green construction projects, according to RSMeans -or on average 12.5% and 20-25% for traditional and building information modeling (BIM)-based 9 projects, respectively, according to the Royal Architectural Institute of Canada 10 ; reduction of uncertainty in cost estimation of construction projects through increased maturity of the engineering design 11 ; expediting the permitting process, which has been found to delay construction by an average of 152.3 days in the developed countries of the Organisation for Economic Co-operation and Development (OECD), according to the World Bank 12 ; providing baseline data and constraints for creating optimal designs, con gurable to the requirements of a new project, particularly by utilizing generative modeling [13][14][15] , and arti cial intelligence (AI)-based evolutionary optimization processes [16][17][18][19][20] ; enabling smaller architectural and consulting rms 21,22 -by employing effective digital transformation frameworks and competitive computational design practices-to design, make decisions (e.g., in procurement and supply chain 23 ), and manage complex and larger projects, and consequently increase their competitive advantage in the market.…”

Machine learning-based inverse design methods considering data characteristics and design space size in materials design and manufacturing: a review

“…While there are existing reviews on ML-based inverse design in materials and manufacturing, 7,10,[53][54][55][56][57][58] few comprehensively discuss the suitability of different methodologies given problem-specific characteristics. This review addresses that gap, offering guidelines for selecting appropriate ML methodologies, considering factors like the scale of design space and data fidelity.…”

“…This approach has been successfully applied to a variety of engineering problems. 55,[121][122][123] Moreover, if low-fidelity data is used to incorporate information from analytical functions (e.g., partial differentiation equations), this concept aligns with the idea of physics-informed neural networks. 124,125 In addition, the concept of transfer learning can also be applied to handle multi-fidelity datasets.…”

Machine Learning In Design Engineering and Manufacturing

“…While MDO frameworks commonly incorporate influences from neighboring stages such as finance [24,25] and production planning [21,25], it is noteworthy that traditional MDO frameworks often overlook the invaluable insights from previous iterations, including manufacturing, quality control, and customer feedback. In addressing this gap, ML emerges as a fitting approach to extract and seamlessly embed this wealth of knowledge into the design process [26].…”