Marble powder is one of the Supplementary Cementitious Materials (SCMs) that used to replace cement for more sustainable concrete, while compressive and Tensile concrete tests are important for us to know the quality of the concrete that we use to build, the testing samples process itself takes at least 28 days for concrete to have its maximum strength to have more reliable results.

The construction industry requires, for sustainability, more natural building materials; the CO₂ emissions are tremendously high, affecting the environment due to the manufacturing of cement (Abbas and Muntean, 2025). The effective solution for the problems considered above will consist in the search for methodologies that minimize excessive consumption of natural materials. Among them, one may list the rational application of industrial waste during the preparation of concrete mixes (Abbas and Muntean, 2025; González-Vallejo et al., 2020; Woź niak et al., 2022; Agarwal and Gulati, 2006; Latawiec et al., 2018). Marble industry generates substantial amounts of waste that can be utilized as a by-product along with cement to produce environmentally friendly concrete (Molnar and Manea, 2016).

Construction waste management (CWM) is a complex adaptive system and a holistic strategy that involves controlled treatment, reduction, reuse, and recycling (Jueyendah et al., 2021), and effective final disposal of waste generated from building, converting waste from the building and manufacturing sectors into valuable products (Trtnik et al., 2009). Simultaneously, it promotes sustainability and cost reduction. Successful CWM reduces damage through avoidance of soil and water contamination, pollution of the atmosphere, and habitat disruption, and avoidance of methane gas emitted by landfill sites (Kelechi Enobie et al., 2024), Through diversion of waste and employment of recycled materials, natural materials and energy are conserved based on the principle of circular economy, where material is kept in circulation over a considerable amount of time (Kelechi Enobie et al., 2024).

A Circular Economy (CE) principle in CWM represents a paradigm shift from a linear “take-make-consume-dispose” model of a conventional economy to a restorative one. The circular process aims to have assets in service for longer, getting the highest value from them with minimal waste generation (Gherman et al., 2023; E. Parliament, 2023). The paradigm is termed a core new sustainable principle to optimize waste management in the construction industry. The advantages of embracing a CE in CWM are. It reduces the amount of waste to landfill sites, saves on natural resources, and even reduces greenhouse gases through reduced new production energy demands. The methodology fosters a more robust building industry that is not so dependent on volatile global supply chains and subject to material shortages.

Blockchain (BC) Technology is one of the best methods that can be used in Waste Management (WM) permanently because BC keeps data inside a chained sequence of blocks, thereby making it simple to track changes through calculating a single-block’s hash value and comparing it with recorded hash value inside a nearby block (Motasem 2024) Researchers proceeded to elaborate on how BC today is used in WM in payment facilitation, tracking, waste monitoring. There are also BC applications that are already in place like Swachhcoin (BC mechanism for effective and eco-friendly micromanagement of trash from homes and businesses and making them into valuable commodities), Recereum (BC platform for recycling wastes and turning them into real value), and Plastic Bank (BC program that focuses on monetizing users in order to keep trash out of ocean). Furthermore, BC can also be used in a cloud system to give a boost to the security of WM making it possible for individuals to participate actively within WM through mobile phone and personal computer (Taylor et al., 2020; Researchgate, 2019). BC even has solutions to core areas named, fraud and manipulation, incorrect/loss of data, manual processing and absence of knowledge and control found within WM practices (França et al., 2020).

In CWM, blockchain integration has a lot of potential by allowing harmonious platforms for managing data and accelerating circular economy CE concepts by means of improved collaboration. That potential can be boosted by drawing on integration strategy applicable to construction. A key strategy involves using smart contracts on platforms like Ethereum to achieve a secure peer-to-peer network within waste processing streams. While references provided do not directly refer to blockchain-based solutions like Swachhcoin, Recereum, Plastic Bank, or DanKu protocol specially applied to waste within construction, the references lay emphasis on blockchain nature like decentralization, immutability, and openness to play a vital role in effective waste tracking and measurement optimization, crucial to CE methods. In addition to that, blockchain’s ability to integrate other digital technologies like BIM, IoT, and RFID, explored to apply to the general context of construction, also supports its contribution to overall waste handling by simplifying real-time tracking, furthering improved data exchange, and automated verifications along waste lifecycle (Mahmudnia et al., 2022; Perera et al., 2020; Yang et al., 2020; Souza Barreto et al., 2020).

By applying ANN (see Figure 1) on the prediction of compressive and tensile strength: This will reduce the amount of time wasted in procedures. It will eliminate such long curing and physical testing activities by directly assessing the quality of concrete. With a well-trained ANN model, decisions can be made quickly over mix design optimization in real time to improve project efficiencies.

The current study aims to provide an ANN model that would be able to predict the compressive and tensile strength of marble concrete. In this respect, 629 data from previous works have been selected with much care, and factors like cement, coarse and fine particles, water, and specimen curing age were taken as a number of these parameters is considered significant determinants of compressive and tensile strength of conventional concrete.

ANN systems have lately become popular, and many researchers have used them for different engineering applications. Laboratory tests are carried out to get the physical and mechanical properties of concrete. By simulating human brains, ANNs may learn from experience the connections between certain inputs and outputs (Molnar and Manea, 2016; Bilim et al., 2009; Demir, 2008; Topc¸u and Sarıdemir, 2008a; Sarıdemir et al., 2009; Lee, 2003; Parichatprecha and Nimityongskul, 2009; O¨ zcan et al., 2009; Topc¸u and Sarıdemir, 2008b; Tanyildizi, 2009). An ANN uses an activation function to generate an output, a learning algorithm to modify the weights, and a set of patterns from a training database as input (Cevik and Guzelbey, 2008; Hou et al., 2008; Topc¸u and Sarıdemir, 2008c).

The ANN is a biologically inspired computational technique that will map the relations of n-dimensional input and output vectors very accurately. In this architecture, the so-called hidden layers composed of hidden neurons are sandwiched between an input and an output layer, where the input layer represents nodes for the input variables and the output layer is utilized to express the outputs of the model. How these three layers are combined is demonstrated in Table 1 (Zadeh, 1965).

These are the connections between the three layers, linked to weights initially set by the network creator but changed iteratively for every “epoch” that the network goes through. Because it reduces the error function of the network output to a predetermined threshold or limit-as instructed by the gradient descent optimization-it provides weight after weight which, in the end, converges with respect to the final values. Or in other words, the number of converged epochs reaches to its maximum until convergence is attained. Each hidden neuron has an associated activation function acting on the weighted sum of inputs it gets from the preceding layer for feeding the result of this function to the subsequent layer and so on until the final output layer. This procedure from the input layer to the output layer is called the feed-forward step. The Backpropagation technique helps in re-adjusting the weights of the network during the learn phase such that the input gets mapped with a more accurate output with a reduced Mean Squared Error. In the process, this results in converging the error to a reasonable margin below a certain threshold (Zadeh, 1965).

It is this mapping capability of ANNs in complex nonlinear relations that have made them quite applicable in fields like engineering and material sciences, including concrete property predictions. While the number of hidden layers and the number of neurons within each layer are problem-dependent, more complicated problems are modeled with deeper architectures. The model is very sensitive to network performance for proper preprocessing of the data, which must at least involve one step of normalization or standardization, in a way that it could make sure the input features must be well scaled to avoid over-problems within exploding or vanishing phases during backpropagation, including diverse activation functions that will be leveraged for capturing complex conditions, Recent breakthroughs in machine learning add the additional capabilities of accurately predicting concrete characteristics, thereby optimizing material performance by reducing the need for excessive physical tests (Trtnik et al., 2009).

Overall, the training of the ANN should be performed on a partition of data into training, validation, and testing subsets so as not to have any generalization problem or overfitting. The model generalization might get improved by applying some regularization techniques, including dropout, or L2 regularization methods. Generally speaking, more sophisticated optimization algorithms, such as the adaptive methods Adam or RMSprop, should be used, since they ensure faster convergence and improvement in general performance. ANNs are flexible in that they can fit into a wide range of problem domains by changing their architecture and hyperparameters, making them a mighty tool for predictive modeling and pattern recognition.

Recent research has explored the use of artificial intelligence to predict the mechanical properties of various sustainable concretes. For example, Jiang et al. (2023) applied AI to forecast the properties of banana peel-ash and bagasse blended geopolymer concrete, demonstrating the versatility of machine learning in sustainable materials. Other studies have proposed simplified AI-based methodologies for geopolymer concrete mix design (Alaneme et al., 2024a), and critical reviews have highlighted the growing role of AI in optimizing geopolymer concrete production (Alaneme et al., 2024b). Additionally, systematic reviews emphasize the eco-friendly potential of agro-waste-based geopolymer concrete (Alaneme et al., 2023a). Compared to these works, this study integrates marble powder as a cement replacement and combines ANN prediction with blockchain-based material tracking, addressing both performance optimization and supply chain transparency.

Though previous work on ANN-based concrete strength prediction exists, few have brought together AI-based prediction and blockchain technology for tracking and sustainability in material sources and construction. This paper presents a novel approach by combining ANN with Blockchain-Rock to predict the mechanical characteristics of marble powder concrete as well as providing transparency and security in material sources and construction. This dual approach encourages real-time mix design decision-making and addresses sustainability challenges in construction.

Future megatrends, that is, urbanisation, resource efficiency, as well as globalisation, are transformative drivers with far-reaching social, economic, business, culture, as well as human impacts. Specifically, digital technologies are remaking these areas in all industries as well as on all continents. Work methods (see Table 2) in design as well as construction have been changed by digitalization (see Figure 2) with highly accurate models that allow planning, designing, constructing, assembling, operation, as well as maintenance (Papayianni and Pachta, 2017).

However, construction material development as well as construction methods have not followed that of architecture and civilizations. All construction technologies developments have in turn relied on construction material (Rajesh et al., 2024), although its sustainability is not always prioritized. The impact buildings have on nature is not realized by society (Papayianni et al., 2016) as much as it is supposed to be. In consideration of digitalization as well as sustainability, construction material as well as construction technologies increasingly have to be reevaluated with consideration towards a more healthy environment.

Several innovations have been made known, and scholars from a variety of disciplines are enthusiastic. Some are developments in deep learning, having big data available for benchmarks, as well as algorithmic developments in terms of activation functions, initialization schemes in weights, as well as improved schemes in terms of optimization. To sense damage in structures (Selvasofia et al., 2020) (Kanhar et al., 2021), cracking in concrete (Shukla and Gupta, 2019) (Sharma et al., 2023), structural elements in bridges (Mishra et al., 2013), in asphalt (Ali and Hashmi, 2014), tunnels (Kumar and Kumar, 2015), towers in transmission (Vaidevi, 2013), as well as ceiling (Zhang et al., 2020), civil engineers have utilized new computer science. The construction sector is far from leading in terms of innovation as a sector because it lacks creativity as well as innovation, as well as more importantly, poor sustainability performances (Chavhan and Bhole, 2014). Innovations involve a broad spectrum of transformative systems such as AI as well as ANN, which can assist in the growth of the construction sector, but whose prospect is eroded by a lack of internet-based data.

Use of marble powder as cement substitute is critical in light of environmental impacts from dumping marble waste as well as construction sector sustainability objectives construction sector is increasingly interested in sustainability. In cement, marble powder, a high surface area material that is a by-product from marble grinding, can be a cement substitute, application of marble powder as cement substitute holds promise in terms of cost as well as environmental impacts in conformity with sustainable construction (Khodabakhshian et al., 2018; Khan et al., 2022; Nega et al., 2023).

629 entries (see Figure 3 and Table 3) (Shukla and Gupta, 2019; Sharma et al., 2023; Mishra et al., 2013; Ali and Hashmi, 2014; Kumar and Kumar, 2015; Vaidevi, 2013; Zhang et al., 2020; Chavhan and Bhole, 2014; Ofuyatan et al., 2019; Ghani et al., 2020; Lezzerini et al., 2022; Chandrasekaran, 2017; Gopi et al., 2017; Rao, 2016; Dhanalakshmi et al., 2022; AshaLak et al., 2019; Chandrakar and Singh, 2017; Majeed et al., 2021; Sharma et al., 2019; Shirule et al., 2012; Pal et al., 2016; Singh et al., 2019; Ris et al., 2014; Sounthararajan and Sivakumar, 2013b; Latha et al., 2015) from available studies on Marble powder (MP) effects on cement as well as durability in concrete have been utilized in this research. Waste management. To promote not just the mechanical properties but also sustainability in concrete, cement in varying ratios replaced these. The output parameter in the database was concrete strength, with input parameters consisting of cement quantity, quantity of fine aggregate (FA), quantity of coarse aggregate (CA), water, Marble powder substitute ratios as well as age in curing.

It is represented that the ANN models based on I-ANN and II-ANN architecture are trained and tested concerning two metrics-Mean Squared Error and correlation coefficient see Table 5 using 60 hidden neurons which resulted in minimum MSE with maximum predictive accuracy.

Results have shown that the addition of Marble powder increases the Compressive and tensile strength of concrete and generally enhances the durability of concrete. The curing time and water-to-cement ratio were the critical factors in controlling the compressive and tensile strength. ANN performance with testing data turned out to be excellent; hence, this approach is applicable (see Figures 6–10 and Table 6).

For better usability, an application model was further developed upon the trained ANN. It further allows users to input certain values: marble powder in percentage, cement, time of curing, and finally water to cement ratio against which concrete compressive and tensile strength is predictable. Figure 11 presents the tool providing immediate insight so as to optimize concrete mixing designs and thereby reduce the current dependence on time-consuming, conventional test methods.

This study developed and validated an artificial neural network (ANN) model for predicting the compressive and tensile strength of marble powder concrete, using a dataset of 629 samples. The model demonstrated high predictive accuracy, outperforming other machine learning techniques. In addition, the study introduced Blockchain-Rock technology to record and verify data related to material sources, mix designs, and predicted properties, thereby enhancing transparency and traceability in the concrete supply chain.

The integration of ANN and Blockchain-Rock offers a comprehensive solution for sustainable construction. ANN enables rapid and reliable prediction of concrete properties, reducing reliance on lengthy laboratory tests, while Blockchain-Rock ensures secure and tamper-proof documentation of all relevant data. This dual approach improves efficiency, quality assurance, and sustainability in the construction industry. Future research may explore real-time IoT integration and expansion to larger datasets to further enhance predictive accuracy and practical applicability.

By leveraging ANN, construction professionals can make real-time, data-driven decisions that improve material performance and structural reliability. The incorporation of marble powder as a supplementary cementitious material (SCM) further aligns with global sustainability efforts by minimizing waste and reducing cement consumption. Additionally, Blockchain-Rock enhances trust and accountability in material sourcing, reducing environmental and economic risks, for further research approaches:

• Expand the dataset to include a wider range of material properties and mix designs.