AI-Driven Optimization in Car

Manufacturing
 CERTIFICATE
CERTIFICATE
This is to certify that the project entitled
“AI-Driven Optimization in Car
Manufacturing ” is a bonafide record of
work done by Saksham Ahuja , Shivay
Gupta , Shivanshi Chakraborty and Aarav
Malhotra in partial fulfilment of the
requirement for award of XII Artificial
intelligence in the CBSE Board during the
academic year 2021- 2022.
AMITY INTERNATIONAL SCHOOL

CLASS XII
2024-25

TEACHERS NAME : Ms. Chavvi

TEAM MEMBERS
1. SAKSHAM AHUJA
2. SHIVANSHI CHAKRABORTY
3. SHIVAY GUPTA
4. AARAV MALHOTRA
 Introduction
The project AI-Driven Optimization in Car
Manufacturing focuses on leveraging artificial
intelligence (AI) to enhance efficiency,
precision, and sustainability in the automotive
production process. The application of AI in
manufacturing encompasses various aspects,
such as optimizing supply chain management,
improving quality control through real-time
defect detection, and streamlining production
line scheduling. By integrating machine
learning algorithms and data analytics, this
project aims to minimize resource waste,
reduce production costs, and enhance overall
productivity.
 Role of Students

Students will actively participate in

designing, implementing, and
testing AI-driven solutions tailored
to the challenges in car
manufacturing. Their roles include
conducting research, developing
machine learning models, analyzing
data, and validating outcomes
against industry benchmarks.
 Significance of the Project

This project provides students with hands-on

experience in cutting-edge technologies while
addressing real-world industrial challenges. It
fosters innovation, bridges the gap between
academia and industry, and contributes to the
advancement of sustainable and intelligent
manufacturing systems.
 Team Roles
Team members and their roles
Role Description
Project Leader: Oversees project progress and coordinates
between team members and stakeholders.

Data Expert: Collects and pre-processes data for training

machine learning models.

Designer: Plans the flow and interface of AI-based systems.

Coder/Prototype Builder:Develops the AI prototype using

Python and other tools.

Tester: Ensures the prototype functions correctly and

meets user requirements.

Project leader: Saksham Ahuja

Data expert: Shivanshi and

Aarav

Designer : Shivay

Coder and prototype builder :

Saksham and Shivay

Tester : Saksham
 Project plan
Phase Task Planned : 20 May
start date Planned : 25 May
End planned: 30 November
Actual start : 4 June
Actual end date : 10 December
duration : 6 months
 Coursework
04-06-2024 to 05-06-2024 -> 5 hours
10-06-2024 to 12-06-2024-> 5 hours( Whole group readings )
Set up a team folder on a shared drive
13-06-2024 ->30 minutes -> Saksham defining the problem
Background
15-06- -2024 to 18-06-2024 ->25 hours
Whole group reading
20-06-2024 to 24-06-2024 -> 25 hours
Research
2-07-2024 to 5-07-2024 -> 20 hours
Whole group issues in our community
6-07-2024 to 10-07-2024 -> 25 hours
Complete section 3 of the Project Logbook
15-07-2024 to 20-07-2024 -> 30 hours
20-07-2024 to 24 -07-2024 ->5 hours (shivanshi and Aarav)
Understanding the users
30-07-2024 to 5 -08-2024 -> 30 hours (shivanshi and Aarav)
Identify users
6-08-2024->30 mins
Interview
10-08-2024-> 30 mins (shivay)
Interview with user (2)
12-08-2024-> 30 minsutes (Aarav)
Complete section 4 of the Project Logbook
15-08-2024 to 30-08-2024-> 45 hours
Brainstorming Team
meeting to generate ideas for a solution
5-09-2024 to 10-09-2024-> 15 hours
15-09-2024 to 23-09-2024-> 25 hours
Complete section 5 of the Project Logbook
25-09-2024-> 2 hours (Saksham)
Designing your solution
2-10-2024 to 5-10-2024-> 15 hours
Whole group meeting to design the solution
6-10-2024->5 hours
Collecting and preparing data
10-10-2024-20-10-2024 -> 25 hours(Shivanshi and Aarav)
discuss data requirements
25-10-2024_>5 hours
Data preparation and labeling
2-11-2024 to 5-11-2024-> 20 hours
Team meeting to plan prototyping phase
7-11-2024->5 hours
Prototyping Testing
Train your model with input dataset
11-11-2024-> 25 hours (Saksham)
Test your model and keep training with more data until you think your model is accurate
12-11-2024 to 17-11-2024-> 20 hours
Team meeting to discuss testing plan
19-11-2024 to 22-11-2024-> 25 hours
Testing Creating the video
Invite users to test your prototype
25-11-2024
Conduct testing with users
27-11-2024
29-11-2024
Team meeting to discuss video creation
30-11-2024
Film your video
4-12-2024 to 6-12-2024 -> 10 hours
Edit your video
7-12-2024
Completing the logbook
Review your project
8-12-2024 to 10-12-2024-> 10 hours
 Communications plan

Communication mode :
Mixture of both offline and online

Frequency of meetings : Twice in a

month
Person responsible for setting up
online document and other :
Team leader.

Tools for communication : Google

meet, Microsoft Teams, Whatsapp
group, zoom.
 Problem Definition
Important local issues faced by school or
community

1. Environmental Sustainability
Local Issue: Pollution from industrial activities
or vehicle emissions.

Solution: AI helps optimize manufacturing

processes to reduce waste and energy
consumption. This can be achieved through
predictive maintenance of machinery, better
resource allocation, and minimizing defects in
production lines, leading to a greener
manufacturing process.

Impact on Community: Reduced emissions and

a cleaner environment benefit the local
population, particularly in areas near factories.
2. Economic Opportunities
Local Issue: Limited job opportunities or
underemployment in the community.

Solution: AI-driven manufacturing can

stimulate economic growth by attracting tech-
driven industries. While automation may
reduce manual labor in some areas, it creates
demand for skilled workers in AI development,
robotics, and maintenance.
Impact on Community: New training programs
and job opportunities for students and local
residents in AI-related fields can improve
employment rates.
3. Traffic and Transportation Issues

Local Issue: Congestion and poor

infrastructure planning.

Solution: AI optimization in car

manufacturing supports the
development of smart and efficient
vehicles, such as autonomous cars and
electric vehicles (EVs), which are better
suited for modern urban environments.
Impact on Community: Improved traffic
management systems and adoption of
EVs reduce congestion and noise
pollution, creating a better quality of
life.
4. Resource Scarcity

Local Issue: Strain on water and energy

resources due to industrial activity.
Solution: AI models can optimize
resource usage in car manufacturing,
ensuring minimal wastage of energy,
water, and raw materials. This can
involve better planning of supply
chains and energy-efficient operations.
Impact on Community: Preservation of
local resources benefits both the
environment and community
sustainability.
. 5. Education and Skill Development

Local Issue: Lack of technical education

programs tailored to modern industry
demands.
Solution: Partnerships between car
manufacturers using AI and local
schools or colleges can create specialized
courses in AI, robotics, and sustainable
manufacturing practices.
Impact on Community: Students gain
hands-on experience and skills relevant
to cutting-edge technology, preparing
them for high-demand careers.
 Focused Issue
How can we use AI to reduce defects and
improve efficiency in car manufacturing
processes?

Car manufacturing is a complex process

involving numerous steps, from raw
material procurement to assembly and
quality testing. Defects and inefficiencies at
any stage can lead to increased costs,
production delays, and compromised
product quality. These challenges highlight
the need for advanced tools and
technologies, such as Artificial Intelligence
(AI), to optimize operations and improve
outcomes.
 The Users
Who Are the Users?

Factory Managers and Operators

In the context of AI-driven optimization in
car manufacturing, factory managers and
operators are the primary users who
directly interact with AI tools and benefit
from their implementation. Their roles are
pivotal in ensuring the smooth operation of
the manufacturing process, improving
efficiency, reducing defects, and
maintaining high-quality standards.
 Problems faced by Users
1. Difficulty in Identifying Defects
Quickly and Accurately

Factory managers and operators often rely on

manual inspections to detect defects, which
can be time-consuming and prone to human
error.

As production speeds increase, it becomes

harder to spot imperfections early in the
process.

Delayed identification of defects may lead to

more significant quality issues down the line,
resulting in waste and costly rework.
2. Challenges in Scheduling Maintenance

Unpredictable machinery breakdowns are a

significant challenge for factory managers.

Scheduling maintenance effectively is

difficult because of limited visibility into the
health of machines.

Reactive maintenance practices often lead to

unplanned downtime, which disrupts
production schedules and increases
operational costs.
3. Lack of Real-Time Insights into
Production Line Performance

Without real-time data, factory

managers cannot respond promptly to
inefficiencies or bottlenecks.

Traditional methods of tracking

performance often rely on manual
reporting, which is time-lagged and lacks
the granularity needed to make
informed decisions quickly.

This lack of visibility prevents them

from optimizing the production process
in real time, leading to reduced
efficiency and potential delays.
 Interview Questions and
User’s Responses
1.What are the highs and lows of their
experience?

Factory managers and operators experience

the benefits of improved production efficiency
and product quality with AI-driven systems
. However, they struggle with the integration
of AI technologies, requiring substantial
training and adaptation to new workflows.

Additionally, the time spent calibrating AI

systems can sometimes interrupt production
schedules.
 2. What are their needs?
They need an efficient AI-driven system
that can identify defects in real-time,
predict maintenance needs, and optimize
production schedules to reduce downtime
and improve overall factory output.

The system should seamlessly integrate with

existing machinery and processes.

3. What's expected of them?

Factory managers and operators are
expected to ensure that the production line
runs smoothly, meet quality standards, and
minimize production delays.
They are also expected to adapt to new
technologies and ensure that AI systems are
effectively utilized to improve
manufacturing efficiency.
 4. who do they work with?

Factory managers and operators work closely

with production line workers, quality control
teams, maintenance personnel, and IT staff.

They must coordinate with various departments

to ensure that AI technologies are implemented
successfully.

5. What do they rely on?

They rely on manual processes for quality

inspection, maintenance scheduling, and
workflow optimization, which can be slow and
prone to human error. T

his reliance often leads to inefficiencies and

higher operational costs.
 6. How do they define success?

Success is defined by the ability to

meet production targets, ensure high
product quality, and minimize
downtime.

They aim for a smooth and efficient

manufacturing process that delivers
consistent quality while reducing
operational costs.
 Empathy Map
What our users are saying:

Factory managers and operators often

express frustration with the time-consuming
nature of quality inspections and
maintenance scheduling. They struggle with
the lack of real-time insights into machine
health and production line performance.

What our users are thinking:

They are considering ways in which AI can

automate defect detection, predict
maintenance issues, and provide real-time
data on production efficiency to improve
overall operations. They are looking for
solutions to minimize downtime and increase
throughput.
 What our users are doing:

They are manually inspecting

components for defects, keeping
detailed records of maintenance
schedules, and relying on outdated
methods for tracking production
line performance. These processes
can be error-prone and slow,
impacting overall efficiency.
 Difficulties Faced by Users

Manual Defect Detection:

The need for manual inspection of components and

products, which is time-consuming and prone to
human error.

Time-Consuming Maintenance Scheduling:

Difficulty in scheduling and tracking maintenance

tasks, often leading to unplanned downtime.

Lack of Real-Time Production Insights:

Without real-time data, managers cannot optimize

workflows or identify inefficiencies immediately,
leading to delays and bottlenecks.

Risk of Human Error in Inspections:

Manual inspections can miss defects or make incorrect

assessments, leading to product quality issues.
 Problem Statement

Factory managers and operators face significant

challenges with the current manufacturing
process due to inefficiencies in defect detection,
unpredictable maintenance schedules, and lack
of real-time insights. This results in increased
downtime, higher operational costs, and
inconsistent product quality, all of which hinder
overall manufacturing performance and
productivity.
 Brainstorming
Ideas
1. AI-Based Defect Detection Using Computer
Vision

AI-powered computer vision systems are

integrated with high-resolution cameras and
sensors to monitor the production line in real
time.

These systems use deep learning algorithms to

automatically identify and categorize defects,
such as cracks, scratches, or misalignments,
with high precision.

By detecting defects at early stages, these

systems reduce waste, improve product quality,
and minimize costly rework.
 2. Predictive Maintenance Using IoT and AI
Models

IoT sensors collect real-time data on machine

performance, such as temperature, vibration, and
pressure, which is then analyzed by AI models.

These models predict when a machine is likely to

fail or require maintenance, allowing operators to
schedule repairs proactively.

This approach reduces unplanned downtime,

extends the lifespan of equipment, and ensures
smoother, more reliable production processes
.
3. Dynamic Scheduling to Optimize Workflow

AI algorithms can analyze real-time production data

to dynamically adjust schedules based on current
conditions, such as machine availability, resource
utilization, and staffing levels.
 Best Fit for Users
AI-based defect detection is the best fit for factory
managers and operators because it significantly
enhances production quality by automating the
identification of defects in real time.

With the ability to quickly detect even the most subtle

imperfections that might be missed by human
inspectors, it helps maintain high-quality standards
and reduces waste.

This automation frees up operators to focus on more

complex tasks, improves consistency, and speeds up
the inspection process, leading to faster production
times.

Ultimately, it leads to cost savings and a more reliable

product, aligning with the operational goals of
improving efficiency and quality.
 Design
Steps to Implement the AI Solution

1. Capture real-time images from the

production line.

2. Pre-process images to enhance clarity.

3. Use a trained computer vision model to

detect defects.

4. Send automated alerts for detected

issues.