A

PROJECT WORK PHASE-I

PROJECT PROGRESS SEMINAR-
I ON
“-”
submitted for particular fulfillment of the requirements of award of degree
In
BACHELOR OF TECHNOLOGY
In
CIVIL ENGINEERING
OF
Rashtrasant Tukdoji Maharaj Nagpur University, Nagpur
Submitted By
Student of 7th semester
Group No. 9
1. Sushma D. Lokhande (109)
2. Tanisha R. Ramteke (111)
3. Hemant R. Wanve (119)
4. Kshitij V. Naranaware(123)

Under Guidance of
Ms. Bharti R.
Gautam [Link].

Department of Civil Engineering

Priyadarshini College of Engineering Nagpur – 440019 (India)
Academic Session 2025-2026

1|Page
 INDEX

[Link]. Content Page No.

1. Introduction 3
2. Aim And Objective 4
3. Detailed Literature Review 5-8
4. Application Of Proposed Project To Society 9
5. Equipment/Materials To Be Required For Project 10-11
6. Methodology 12-13
7. References 14-16

2|Page
 1. INTRODUCTION
 Concrete is one of the most widely used construction materials in the
world, serving as the backbone of infrastructure, buildings, and other civil
engineering projects. Its compressive strength is the most crucial
mechanical property determining its structural performance, durability,
and safety. Traditionally, compressive strength is evaluated through
destructive testing methods, such as casting and curing specimens
followed by laboratory compression tests. While effective, these methods
are time- consuming, labor-intensive, and not always feasible during the
construction process.
 With advancements in computational techniques, machine learning
approaches—particularly Artificial Neural Networks (ANNs)—have
emerged as powerful tools for predicting material properties, including the
compressive strength of concrete. ANNs mimic the human brain’s ability
to learn and generalize from data, making them highly effective in
capturing the complex, nonlinear relationships between various mix
components and strength outcomes.
 This report explores the application of ANN models in predicting the
compressive strength of concrete. The document includes a literature
review of previous studies, outlines aims and objectives, discusses
methodology, evaluates the societal application of the proposed project,
and provides conclusions and recommendations for future development.
 A small-scale ANN model is trained on experimental data for strength
forecasting. The ANN can predict strength quickly without actual testing,
which helps engineers save time and reduce costs. It also improves
decision- making in concrete mix design and quality control.
 The study includes a literature review of past research and real-world case
studies. The main aim is to develop and evaluate an ANN model that can
accurately predict the compressive strength of concrete based on mix
parameters

3|Page
 2. AIM AND OBJECTIVES
Aim :-

To develop and evaluate an Artificial Neural Network (ANN) model that can
accurately predict the compressive strength of concrete by NDT.

Objectives :-
To develop and evaluate an Artificial Neural Network (ANN) model that
can accurately predict the compressive strength of concrete using only
non- destructive test (NDT) data.
1. To collect concrete quality data through non-destructive testing
methods such as Rebound Hammer Test and Ultrasonic Pulse Velocity
Test.

2. To preprocess and normalize the NDT data for effective ANN training.

3. To design and train an ANN model with NDT parameters as input.

4. To compare the ANN predictions with available reference strength

values to assess accuracy and reliability.
5. To demonstrate the potential of ANN combined with NDT as a fast,
cost- effective, and non-invasive tool for predicting concrete strength in
field applications.

4|Page
 DETAILED LITERATURE REVIEW

[Link] Author(s) and Title abstract

. Year
1. Mario Bonagura, Artificial Neural Artificial Neural Networks
Lucio Nobile Network (ANN) (ANNs) for predicting concrete
(2021) Approach for compressive strength using SonReb
Predicting non-destructive tests (rebound
Concrete hammer + ultrasonic pulse
Compressive velocity). The results show that
Strength by ANNs outperform traditional
SonReb correlation formulas, providing
more accurate estimates compared
to core test measurements
2. Yeh, I.C. (2018) Artificial Used ANN with 1030 datasets
Neural including cement, slag, fly ash,
Network Model water, superplasticizer, coarse and
for High- fine aggregate, and curing age.
Performance ANN showed better prediction
Concrete Strength than statistical models,
Prediction highlighting w/c ratio and cement
as most influential parameters.
construction material.
3. Vaseilios [Link], Concrete This study uses artificial neural
Panagiotis G. compressive networks (ANNs) to predict
Asteris strength using concrete compressive strength
(2019) artificial neural from non-destructive tests like
networks ultrasonic pulse velocity and
Schmidt hammer. ANNs provide
reliable estimates closely
matching experimental results,
and the model weights are shared
for easy practical implementation
4. Tanyildizi, H. & Prediction of Designed ANN using experimental
Çevik, A. (2017) Concrete Strength concrete data and validated
Using ANN and predictions with lab results. ANN
Experimental predictions closely matched
Validation experimental compressive strength
values, proving reliability for
practical use in building projects.

5|Page
5. Khademi, Application of ANN Investigated ANN performance in
F. et al. (2016) in Forecasting predicting strength of eco-friendly
Compressive concrete mixes with fly ash.
Strength of Results indicated ANN can save
Sustainable time and cost in designing
Concrete sustainable concrete without
extensive lab trials.

6. Khan et al. Application of ANN created an ANN model trained on

(2023) in Predicting a large dataset of normal and
Strength of High- high- strength concrete. They used
Volume Fly Ash eight input parameters including
Concrete cement, water, aggregates, slag,
fly ash, superplasticizer, and
curing age. Their optimized ANN
of Composite Board achieved very good accuracy (R²
for Ceiling Tiles ≈ 0.93) and showed that cement
from Sawdust, content and superplasticizer were
Waste Paper, and the most important factors
PET Plastic affecting compressive strength.

7. Lee, S.C. (2003) Modeling the Developed an ANN with back-

Compressive propagation to predict compressive
Strength of Concrete strength of various concrete
Using ANN mixes. Demonstrated that ANN
significantly reduces the need for
laboratory experiments by
accurately predicting strengths at
different ages.
8. Sobhani ANN-Based Used ANN to predict strength of
, J. et al. (2010) Prediction of concrete containing fly ash, slag,
Compressive and silica fume. Showed that
Strength of Concrete ANN captures complex
Containing interactions between SCMs and
Supplementary base materials, achieving high
Cementitious accuracy compared to
Materials conventional methods.

6|Page
9. Arslan Qayyum Optimized artificial An ANN model using Levenberg-
Khan, Hasnain neural network Marquardt Backpropagation
Ahmad Awan, (ANN)model for accurately predicted concrete
Mehboob Rasul, accurate prediction compressive strength from 1637
Zahid Ahmad of compressive samples, with cement and
superplasticizer being the most
Siddiqi, Amorn strength of normal
influential factors.
Pimanmas and high strength
(20 concrete

10. Murts et al. Predicting the An ANN model using six input
(2021) compressive parameters accurately predicted the
strength of ground compressive strength of GGBFS
granulated blast concrete across 45 mixes and ages
furnace slag up to 360 days, confirming its
concrete using effectiveness as an alternative to
artificial neural experimental testing.
network

11. Sherin Khadeeja Structural ANN models can accurately predict

Rahman, assessment of Basalt the flexural strength of Basalt FRP-
Riyadh Al- FRP reinforced self- reinforced geopolymer concrete
Ameri(2023) compacting beams under ambient and marine
geopolymer concrete exposures, outperforming existing
empirical and numerical models
using artificial
neural network
(ANN) modelling

12. Yinglong Wu , This study developed an optimized

Optimizing pervious
R. Pieralisi , F. MLP-based machine learning model
concrete with to predict the mechanical and
Gersson B. machine learning:
Sandoval , hydraulic performance of pervious
Predicting concrete, achieving high accuracy
Lopez-Carreno, permeability and
Ricardo (R² = 0.97–0.98) and reducing
compressive strength reliance on trial-and-error
Pieralisi(2024) using artificial neural experiments.
networks

7|Page
13. Gholamreza Proposing an an Adaptive Fine-Tuned Neural
Pazouki, optimised data- Network (AFTNN) that
Nariman efficient artificial outperformed RBFNN in predicting
Saeed, Won- neural network compressive strength of eco-
Hee model for estimating friendly concretes with SCBA and
Kang(2025) the compressive copper tailings, achieving high
strength of concrete accuracy (R ≈ 0.93) even with
containing SCMs limited training data..
14. Jyoti Thapa concrete An optimized ANN model using
compressive strength 776 datasets accurately predicted
prediction by concrete compressive strength (R² =
artificial neural 0.87), showing cement and time as
network key positive influencers, and water
approach and fly ash as negative, highlighting
ANN’s effectiveness for
and
construction applications.
Sawdust
15. Palika Prediction of Concrete compressive strength
Chopra, Compressive prediction using ANN (Levenberg-
Rajendra Strength of Concrete Marquardt) and Genetic
Kumar Using Artificial Programming showed that the
Sharma,Mane Neural Network and ANN model outperformed GP,
providing the most accurate
ek Kumar Genetic
predictions for both lab and in situ
Programming data.
16. Mahfoud Prediction of the An ANN model with 18 inputs
Benzerzour, Compressive accurately predicted the
Rachid Strength of Waste- compressive strength of 1303
Zentar, Nor- Based Concretes waste- based concrete mixtures,
Edine Abriak Using Artificial achieving high precision (R² =
(2022) Neural Network 0.9888, MAPE
= 2.87%) using MATLAB.

8|Page
 3. APPLICATION OF PROPOSED TO SOCIETY

 Quick Strength Forecasting – Predicts compressive strength with

non-destructive testing.
 Efficient Mix Design – Helps in selecting cost-effective and high-
performance concrete mixes.
 Reduced Cost & Time – Minimizes trial mixes and laboratory work.
 Quality Control – Ensures consistent strength in building projects.
 Sustainable Concrete – Accurately models strength with fly ash, slag,
or recycled materials.
 Construction Planning – Assists in deciding formwork removal
and load application timing.
 Structural Safety – Prevents under-strength concrete use, enhancing
safety.
 Smart Building Integration – Can link with IoT/BIM for real-time
monitoring.

9|Page
 4. EQUIPMENT/MATERIALS TO BE REQUIRED
FOR PROJECT
Materials :-

Tools And Equipment :-

 Rebound Hammer – to measure surface hardness of concrete.

 Ultrasonic Pulse Velocity (UPV) Tester – to measure concrete quality and detect internal flaws.

 Cover Meter / Ferroscan (optional) – to locate reinforcement if needed.

Fig. Fig. Weighing Balance

Compression Testing Machine

10 | P a g e
 Fig. Rebound Hammer Fig. ultrasonic pulse velocity

SOFTWARE TOOLS:-
Software Tools – Python (with libraries such as TensorFlow, Keras, Scikit-learn, Pandas, NumPy,
Matplotlib) or MATLAB.
Computer/Laptop – High-performance system with sufficient RAM and processor speed for
ANN computations.

ANN MODEL PREDICTIING THE COMPRESSIVE STRENGTH OF

CONCRETE
11 | P a g e
 5. Methodology

Following Methodology is to be carried out to make composite tiles :-

1. Data Collection: Perform non-destructive tests (Rebound Hammer, UPV) on concrete samples and
record readings.
2. Data Preprocessing: Organize and normalize the collected data for ANN training.
3. ANN Model Design: Select inputs (NDT values) and output (compressive strength); design the ANN
architecture.
4. Training & Validation: Train the ANN model with part of the data and validate it with the remaining
data.
5. Performance Evaluation: Check accuracy using R², RMSE, and MAPE; compare predicted
values with reference data.
6. Analysis & Conclusion: Analyze results and assess the effectiveness of ANN for predicting concrete
strength.

12 | P a g e
 6. METHODOLOGY

Following methodology is to be used :-

13 | P a g e
 7. References

1. Lin, Chia-Ju, and Nan-Jing Wu. "An ANN Model for Predicting the
Compressive Strength of Concrete." Applied Sciences, vol. 11, no. 9, 2021,
p. 3798 . [Link]
2. Khan, A. Q. (2023). Optimized artificial neural network model for accurate
prediction of concrete compressive strength. Materials Today:
Proceedings,77,1846–
[Link]://[Link]/science/article/pii/S2772397623000448
#bi005
3. Rahman, S. K., et al. (2023). Structural assessment of Basalt FRP
reinforced self-compacting geopolymer concrete beams under flexural
loading. Journal of Cleaner
Production, 396, 136436.
[Link]
4. Wu, Y. "Optimizing Pervious Concrete with Machine Learning."
Construction and Building Materials, vol. 396, 2024, p. 137619.
[Link]
5. de Prado-Gil, J., Martínez-García, R., Jagadesh, P., Juan-Valdés, A.,
González-Alonso, M.-I., & Palencia, C. (2024). To determine the
compressive strength of self-compacting recycled aggregate concrete using
artificial neural network (ANN). Ain Shams Engineering Journal, 15(2),
102548. [Link]
5. R., R., and S. Mandal. "Prediction of Compressive Strength of High
Volume Fly Ash Concrete Using Artificial Neural Network." Proceedings
of the International Conference on Sustainable Construction and Building
Materials (SCBM), NITK Surathkal, Mangalore, India, [Link], T.
O., et al. (2021).Polymer tiles from polyethylene terephthalate (PET)
wastes. F1000Research, 10, 113
A. Khan et al., "Optimized artificial neural network model for accurate
prediction of compressive strength of normal and high strength concrete,"
Science Progress, vol. 106, no. 3, p. 003685042311785, 2023.
[Link]
6. Lee, S.C. "Prediction of Concrete Strength Using Artificial Neural
Networks." Engineering Structures, vol. 25, no. 7, 2003, pp. 849–857
[Link] .
7. Pazouki, Gholamreza, Nariman Saeed, Zhong Tao, and Won Hee Kang.
14 | P a g e
 "Proposing an Optimised Data-Efficient Artificial Neural Network Model for
Estimating the Compressive Strength of Concrete Containing SCMs."
Structures, vol. 79, 2025, p. 109531
[Link]
Khan, A., Usman, M., Shahid, M., & Rizwan, M. (2023). Optimized artificial
neural network model for accurate prediction of compressive strength of normal
and high strength concrete. Cleaner Materials, 10, 100211.
[Link]

8. Ramachandra, R., and S. Mandal. "Prediction of Compressive Strength of

High Volume Fly Ash Concrete Using Artificial Neural Network." In
Proceedings of the International Conference on Sustainable Construction
and Building Materials (SCBM), NITK Surathkal, Mangalore, India, 2018–
[Link]
9. Thapa, Jyoti. "Concrete Compressive Strength Prediction by Artificial
Neural Network Approach." Journal of Engineering Issues and Solutions,
vol. 3, no. 1, 2024, pp. 76–90. [Link]
Ramachandra, R., & Mandal, S. (2018). Prediction of compressive strength of
high volume fly ash concrete using artificial neural network. Proceedings of the
International Conference on Sustainable Construction and Building Materials
(SCBM), NITK Surathkal, Mangalore, India.
[Link]
10. A modified firefly algorithm-artificial neural network expert system for
predicting compressive and tensile strength of high-performance concrete
10.1016/[Link].2018.05.201**

11. Application of artificial neural networks to predict compressive

strength of high strength concrete 10.5897/IJPS11.023
12. Bilim, C.; Atiş, C.D.; Tanyildizi, H.; Karahan, O. Predicting the
compressive strength of ground granulated blast furnace slag concrete
using artificial neural network. Adv. Eng. Softw. 2009, 40, 334–340.
[Link]

15 | P a g e
 Presented By Group No. 9 Under Guidance
Of [Link] R.
Gautam

Sushma D. Lokhande (109) [Link]

Tanisha R. Ramteke (111)
Hemant R. Wanve (119)
Kshitij V. Naranaware(123

16 | P a g e