Construction and demolition waste (DW) generation information has been recognized as a tool for providing useful information for waste management. Recently, numerous researchers have actively utilized artificial intelligence technology to establish accurate waste generation information. This study investigated the development of machine learning predictive models that can achieve predictive performance on small datasets composed of categorical variables. To this end, the random forest (RF) and gradient boosting machine (GBM) algorithms were adopted. To develop the models, 690 building datasets were established using data preprocessing and standardization. Hyperparameter tuning was performed to develop the RF and GBM models. The model performances were evaluated using the leave-one-out cross-validation technique. The study demonstrated that, for small datasets comprising mainly categorical variables, the bagging technique (RF) predictions were more stable and accurate than those of the boosting technique (GBM). However, GBM models demonstrated excellent predictive performance in some DW predictive models. Furthermore, the RF and GBM predictive models demonstrated significantly differing performance across different types of DW. Certain RF and GBM models demonstrated relatively low predictive performance. However, the remaining predictive models all demonstrated excellent predictive performance at R² values > 0.6, and R values > 0.8. Such differences are mainly because of the characteristics of features applied to model development; we expect the application of additional features to improve the performance of the predictive models. The 11 DW predictive models developed in this study will be useful for establishing detailed DW management strategies.

Recently, artificial intelligence (AI) technologies have been employed to predict construction and demolition (C&D) waste generation. However, most studies have used machine learning models with continuous data input variables, applying algorithms, such as artificial neural networks, adaptive neuro-fuzzy inference systems, support vector machines, linear regression analysis, decision trees, and genetic algorithms. Therefore, machine learning algorithms may not perform as well when applied to categorical data. This article uses machine learning algorithms to predict C&D waste generation from a dataset, as a way to improve the accuracy of waste management in C&D facilities. These datasets include categorical (e.g., region, building structure, building use, wall material, and roofing material), and continuous data (particularly, gloss floor area), and a random forest (RF) algorithm was used. Results indicate that RF is an adequate machine learning algorithm for a small dataset consisting of categorical data, and even with a small dataset, an adequate prediction model can be developed. Despite the small dataset, the predictive performance according to the demolition waste (DW) type was R (Pearson’s correlation coefficient) = 0.691–0.871, R² (coefficient of determination) = 0.554–0.800, showing stable prediction performance. High prediction performance was observed using three (for mortar), five (for other DW types), or six (for concrete) input variables. This study is significant because the proposed RF model can predict DW generation using a small amount of data. Additionally, it demonstrates the possibility of applying AI to multi-purpose DW management.

Response scales are widely used to assess the personal experience of sensation and perception in built environments, and have a great impact on the quality of the responses. The purpose of this study was to investigate the effects of response scales on human sensation and perception in moderate indoor environments. Four different response scales were compared under three room temperatures (19.0 °C, 24.5 °C, and 30.0 °C) and five acoustic stimuli (ambient noise, 42 and 61 dBA × water sounds and traffic noise): a bipolar seven-point scale according to ISO 10551:1995, a unipolar 11-point scale according to ISO/TS 15666:2003, these two scales combined for each sensory comfort assessment, and a bipolar visual analogue scale. The degree of relative differentiation based on indoor physical factors made no significant difference across the four response scales. Therefore, the effects of physical factors on human response could be assessed by using any of the four scales tested in this study, with a statistical significance at p < 0.05 in moderate environments. The choice of response scale would depend not only on the type of physical stimulus but also on the question of sensation or perception. The reliability of each response scale was different according to the subjective attributes. The bipolar visual analogue scale was subjectively preferred by the respondents.

This study presents a multi-objective parametric design tool for four-axis surround-type movable shading device using solar position tracking in Seoul, South Korea. In order to explore large numbers of possible forms of shades, generic algorithms are utilized with real-time simulation of the performative criteria such as solar radiation, daylight glare probability (DGP), and solar shielding rate on window surface. This study outlines a workflow using a multi-objective engine called Octopus that runs within Grasshopper 3D, a parametric design tool, in addition to environmental performance simulation plug-in Ladybug. The workflow utilizes a performance-based design tool, which allows the designer to explore, sort, and filter solutions, and visually compare alternative solutions in terms of energy saving and indoor daylight quality in order to determine the optimal form of shade changing its shape every one hour. The result of deriving and analyzing the optimal shade shape through the genetic algorithm proposed in this study is as follows: On the one hand, on the summer solstice, shade shapes with shielding areas of almost 100% should be derived to achieve the most effective reduction of the direct solar radiation. The proposed movable shading device reduced direct solar radiation by 52.40% and 57.20% in the south- and east-facing windows, respectively. On the other hand, in winter when solar heat gain is important, the absence of sunshade is optimal in terms of heating load. However, in order to improve the indoor light environment, it is confirmed that it is possible to derive a certain shape of sunshade according to the sun’s trajectory. On the winter solstice, the problem of glare arises from 10:00 to 15:00 in the south and 10:00 in the east. Therefore, the proposed four-axis movable shading device can be configured to have a minimum protrusion length satisfying DGP less than 0.35 in winter.

Maintaining a pleasant indoor environment with low energy consumption is important for healthy and comfortable living in buildings. In previous studies, we proposed the integrated comfort control (ICC) algorithm, which integrates several indoor environmental control devices, including an air conditioner, a ventilation system, and a humidifier. The ICC algorithm is operated by simple on/off control to maintain indoor temperature and relative humidity within a defined comfort range. This simple control method can cause inefficient building operation because it does not reflect the changes in indoor–outdoor environmental conditions and the status of the control devices. To overcome this limitation, we suggest the artificial intelligence integrated comfort control (AI2CC) algorithm using a double deep Q-network(DDQN), which uses a data-driven approach to find the optimal control of several environmental control devices to maintain thermal comfort with low energy consumption. The suggested AI2CC showed a good ability to learn how to operate devices optimally to improve indoor thermal comfort while reducing energy consumption. Compared to the previous approach (ICC), the AI2CC reduced energy consumption by 14.8%, increased the comfort ratio by 6.4%, and decreased the time to reach the comfort zone by 54.1 min.

As people spend more than 90% of their time indoors, indoor environmental quality (IEQ) has become an important factor in maintaining a healthy space for the occupants. There are many indoor climate control devices for improving IEQ. However, it is difficult to maintain an appropriate IEQ with changing outdoor air conditions and occupant behavior in a building. This study proposes a reinforcement learning (RL) algorithm to maintain indoor air quality (IAQ) with low energy consumption in a residential environment by optimally operating indoor climate control devices such as ventilation systems, air purifiers, and kitchen hoods. The proposed artificial intelligence algorithm (AI2C2) employs DDQN (double deep Q-network) to determine the optimal operation of various indoor climate control devices, reflecting IAQ and energy consumption via different outdoor levels of particulate matter. This approach considers the outdoor air condition and occupant activities in training the developed algorithm, which are the most significant factors affecting IEQ and building energy performance. A co-simulation platform using Python and EnergyPlus is applied to train and evaluate the model. As a result, the proposed approach reduced energy consumption and maintained good IAQ. The developed RL algorithm for energy and IAQ showed different performances based on the outdoor PM 2.5 level. The results showed the RL-based control can be more effective when the outdoor PM 2.5 level is higher (or unhealthy) compared to moderate (or healthy) conditions.