Development of AI Model Integrating Experimental Analysis for Predicting the Mechanical Properties of E-Waste concrete
PROJECT SUMMARY: Introduction The exponential development of the electronic industry and changes in people?s lifestyle have increased the discarding rate of waste electronic appliances and electrical equipment?s rapidly. The disposal of electronic waste (e-waste) ha
2025-06-28 16:26:39 - Adil Khan
Development of AI Model Integrating Experimental Analysis for Predicting the Mechanical Properties of E-Waste concrete
Project Area of Specialization Artificial IntelligenceProject SummaryPROJECT SUMMARY:
Introduction
The exponential development of the electronic industry and changes in people’s lifestyle have increased the discarding rate of waste electronic appliances and electrical equipment’s rapidly. The disposal of electronic waste (e-waste) has become a serious challenge to developing and developed countries as toxic substances and heavy metals present in e-waste could harm human health and the atmosphere. From the perspective of environmental concerns, it is necessary to properly dispose or reuse e-waste in any forms to prevent pollution. In recent years, a growing number of studies are primarily associated with utilization of e-waste as construction materials. In E-waste Concrete different types of e-waste can be utilized as source material in the form of binder, fine/coarse aggregate, and fibre in the mortar, concrete, and precast products. Overall, it is suggested that e-waste offers huge potential benefits when it is utilized in construction products, thereby reduces e-waste management problems and saves the earth from environmental pollution.
The Research Problem
In the above discussion we discussed researchers have already worked on e-waste concrete but after analyzing and deep studying we found that an AI model should be developed for E-waste Concrete to predict the mechanical properties of concrete by using new techniques of Artificial Intelligence.
The Purpose of the Study
The mechanical properties will be estimated by the application of proposed simplified mathematical expression. The accuracy, generalization and prediction capability of the proposed model will be assessed by conducting parametric analysis, applying statistical checks, and comparing it with regression models. This study will enhance the re-usage of e-waste for the development of green concrete leading to environmental protection and monetary benefits
PROJECT OBJECTIVES:
Researches is available on the experimental determination of the mechanical properties of e-waste concrete but that is not enough to conclude and no proper mix design procedure is developed. There is a need to accurately formulize the mechanical properties of E-waste Concrete. Keeping these points in mind, the following objectives are formulated;
- To Generate a Data Points by performing an extensive experimental work for E-waste Concrete that will be performed with different combinations and proportions to find its Mechanical Properties which includes Compressive Strength, Split Tensile Strength and Flexural Strength.
- To employ the different machine learning techniques for determining a model that can accurately predict the Compressive Strength, Split Tensile Strength and Flexural Strength of E-waste Concrete.
- To perform a Sensitivity and Parametric Analysis to check for the relative contribution of different selected input variables.
Data Preparation
The data will be obtained from the previous research studies and further extensive experimental work carried. As mentioned for E-waste concrete past study is not done substantially so the data is not enough to prepare an AI Model. So, further extensive experiments should be performed that consider large group of variables and conclude the results from all type of combinations and proportions which would help to complete the dataset for modeling. Input variables namely, (i) %age of E-waste Aggregate replaced as Coarse Aggregate in Concrete, (ii) %age of E-waste Aggregate replaced as Fine Aggregate in Concrete, (iii) Water to Cement ratio, (iv) Age of Specimen, (v) Specific Gravity of E-waste Aggregate, (vi) Specific Gravity of Coarse Aggregate, (vii) Specific Gravity of Fine Aggregate, (viii) Water Absorption of E-waste Aggregate, (ix) Water Absorption of Coarse Aggregate, (x) Water Absorption of Fine Aggregate. Outputs includes (i) Compressive Strength, (ii) Split Tensile Strength and (iii) Flexural Strength.
Model Architecture
Three AI-based models will be employed in this study to predict the mechanical properties of E-waste Concrete. Firstly, the Artificial Neural Networks model will be used. The ANN model simulates the function of the biological nervous systems in the human brain to solve the problems. Basically, a neuron receives input values from neurons on the previous layers, computes an output, and sends the results to all connected neurons on the next layer. Second approach will be Gene Expression Programming technique in which a fix length of chromosomes is used that encodes the program and gives a simple and reliable mathematical equation that can be used practically to predict strength of concrete. Third approach will be based on the technique adaptive neuro fuzzy interface (ANFIS) which optimizes the fuzzy inference system. ANFIS constructs a series of fuzzy if–then rules with appropriate membership functions to produce the stipulated input–output pairs. ANFIS can modify these fuzzy if–then rules and membership functions to minimize the output error measure or explain the input–output relationship of a complex system.
Model Development and Performance
Input Variables mentioned above and the strength of E-waste concrete was assigned as an output. The dataset was divided randomly into 3 subsets, in which 70% of the entire dataset was employed for training model, 15% data for validation, and the remaining 15% was utilized for testing the prediction accuracy of the model.
For ANN several trials will be performed to determine an optimal number of the hidden layers as well as the number of neurons. For GEP optimized no of chromosomes, gene head size, no of constant per gene and mathematical functions used in the equation. The value of R, R2, MAE, RSE, NSE, RMSE, RRMSE, Performance Index, Objective Function presents the ability of a model to predict the outputs based on the inputs.
BENEFITS OF THE PROJECT:
- E-waste have become an integral part of daily life which provides more comfort, security, and ease of exchange of information. These electronic waste (E-Waste) materials have serious human health concerns and require extreme care in its disposal to avoid any adverse impacts. Disposal or dumping of these E-Wastes also causes major issues because it is highly complex to handle and often contains highly toxic chemicals such as lead, cadmium, mercury, beryllium, brominated flame retardants (BFRs), polyvinyl chloride (PVC), and phosphorus compounds. Hence, E-Waste can be incorporated in concrete to make a sustainable environment.
- Concrete is one of the most used materials in buildings today; yet, predicting the accurate concrete strength remains challenging because of the highly complex relationship between its mixture. An accurate method of predicting concrete strength can provide a significant advantage to the construction material industry, particularly within the concrete material industry. Many methods can be used to build the prediction model of concrete strength. However, the traditional methods have so many shortcomings, including expensive experimental costs and the inability to formulate an accurate complex relationship between the components of a concrete mixture with the strength. To overcome this issue, this study applies multiple artificial intelligence (AI) methods to find the most accurate input and output relationships within concrete mixtures.
- The significance of this research is that we can predict the mechanical properties of E-waste Concrete without any destructive experiments by using AI Model. The most valuable significance of AI modeling is less time consuming. Operating AI can also reduce cost which can be spend on experimental procedures.
TECHNICAL DETAILS OF FINAL DELIVERABLES:
In this Final Year Project an AI Model will be prepared based on ANN, GEP and ANFIS techniques that will predict the mechanical properties of E-waste concrete which includes compressive strength, split tensile strength and flexural strength. After modeling statistical analysis, parametric analysis and sensitivity analysis of all the three models for each property of E-waste concrete will be carried out and after comparative study of these models Final Model can be decided based on the performance, analysis results and reliability of the models that can predict the strength of concrete most efficiently. These models include an AI based Equations that include an input of variables and output as strength of E-waste concrete. This equation will be different for different mechanical property of E-waste concrete. Three equations will be finalized based on the most efficient model for compressive strength, split tensile strength and flexural strength each. After finalizing an equation for each property, a graphics integrated user interface software will be made that would include an input in which equations are back programmed to give an output result of E-waste concrete strength (Compressive Strength, Split Tensile Strength and Flexural Strength). This can be done with help of C# language software. This output would be just for finding the mechanical properties of E-waste concrete but for future we have planned to integrate all type of concrete strength in this software. This would be the complete AI based software that can predict the strength of all type of concretes. The Models and equation for other concrete types will be extracted from the most efficient past research studies that worked on the AI models for different concrete. The other type of concrete can be Normal Concrete, Fly-Ash Concrete, Bagasse Ash Concrete, Plastic Concrete, E-waste Fly-Ash Concrete, Geo-Polymer Concrete etc. In Civil Engineering work has been done and is still working on the application of AI in the field of Civil Engineering but it is still not worthy because there is no working on the commercialization of these AI based studies done in field of Civil Engineering. So the next step is to commercialize this software as the only AI based Model that can predict different type of concrete mechanical properties.
| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 79250 | |||
| Material (Cement) for Experimental Work | Equipment | 25 | 750 | 18750 |
| Material (Crush) for Experimental Work | Equipment | 1 | 12000 | 12000 |
| Material (Sand) for Experimental Work | Equipment | 1 | 24000 | 24000 |
| Material (E-waste Aggregate) for Experimental Work | Equipment | 250 | 58 | 14500 |
| Printing Thesis and Panaflex | Miscellaneous | 1 | 3000 | 3000 |
| PVC Pipes for Specimen casting | Miscellaneous | 28 | 250 | 7000 |