Entering the 4th Industrial Revolution: Supply Chain Efficiency in RMG by using AI based Machine Learning Technology

Name: Akib Hasan Srabon Name: Anindita Ishrat
Name: Promita Rani Mondal PewName: Anirudha Sankar Roy
MXUB-0117-19

Entering the 4th Industrial Revolution: Supply Chain Efficiency in RMG by using AI based Machine Learning Technology
Page | 1
Entering the 4th
Industrial Revolution: Supply Chain Efficiency in RMG
by using AI based Machine Learning Technology
According to the IMF, since 2004, Bangladesh averaged a GDP growth of 6.5%, which
has been importantly driven by its exports of readymade garments. Bangladesh’s export
earnings from the apparel sector registered an 8.76% growth reaching $30.61 billion in
FY18. Bangladesh has around 5,000 garment factories, 95% of the garments are owned
by local companies. Around 10 million livelihoods directly or indirectly depend on it.
Considering the current scenario of the RMG sector in Bangladesh, the significance of
Supply Chain Management is considerably ignored and the number of Supply Chain
Management professionals in the apparel industry is really meager.
The foremost shortcoming that is widespread in the middle management of apparel
industry of Bangladesh is the unskilled labor although there exists abundant supply of
labour. The existing challenges will multiply if suppliers aren’t able to fill higher-skill middle
management positions and if the amount of skilled workers needed in Bangladesh’s
apparel industry is not secured.
In a report published by McKinsey in 2011, all types of stakeholders during an interview
mentioned the lack of skilled middle management as a key factor limiting productivity
improvement at suppliers. They added:
 There are very few or nonexistent educational institutions for technical skills.
 The young top employees and graduates are not really fascinated by the current
image of apparel industry.
 Several problems such as cultural issues and the lack of incentives for local
workers to pursue internal training or development persist. As a result, the low
educational level of workers contributes to more in competencies in the garment
factories. Experts estimate that there is currently a 25 percent shortage of skilled
workers in Bangladesh’s apparel industry.

Page | 2
The above mentioned talent crisis has further led to the rise of following challenges and
issues in the supply chain:
 Lack of strategic planning.
 Unavailability of appropriate information leads to incorrect estimation of the
inventory that results in variation among the exact requirements, the order placed
and the delivered inventory, also leading to increased inventory level, risk of stock
obsolescence and hiring of expensive warehouses.
 Absence of easily on-hand middle management
 Lag between customer requirements and supplier capabilities
 Lack of investment in new technologies and machineries
 Business owners and managers lack adequate knowledge about the industry and
base their decisions on profit margins and not future prospects.
 Lack of more out spoken and experienced Merchandisers.
 Poor flow of information within the companies due to the implementation of
traditional management practices and lack of technology.
 Communication gap between the suppliers and the manufacturer due to the
manufacturer’s fear of revealing crucial business information to its competitors.
Note that all of the above challenges and flaws in the supply chain increase lead time
and costs and hamper quality. A simplified model of the current apparel supply chain
is depicted below.
Image: Configuration of Apparel Supply Chain

Page | 3
Therefore, our big idea is
"To reduce the lead time by the application of AI based Machine Learning
technologies in the apparel industry of Bangladesh"
Machine learning is revolutionizing apparel inspections by providing prescriptive insights
into every step of the process and improving quality, reducing risk and increasing
transparency. The greatest challenge to improving the product quality apparel factories
are delivering today is capturing accurate, credible inspection data. With many factories
relying on decades-old methods of recording inspections on paper, in Microsoft Word or
Excel, they’re finding these methods can’t keep up with the pace brands and retailers
need to produce products at.
Factories, brands, retailers and the customers they serve know whether inspections are
succeeding or not by the quality of the garments produced. Inspections make it possible
for a brand or retailer to make a positive impression instantly by providing customers with
the highest quality garments possible. Automating inspections using mobile technologies,
aggregating and analyzing data on a secure, cloud-based inspection platform, and
gaining prescriptive insights from the data using machine learning is revolutionizing
apparel manufacturing.
Capturing real-time data streams from inspections, quickly analyzing and taking
prescriptive actions on the insights gained using machine learning is driving inspections’
inflection point today. By using constraint-based algorithms and logic to understand
why there are large differences in inspection and product quality between factories,
it’s possible to improve an entire supply chains’ product quality and delivery performance.
Machine learning algorithms also “learn” or optimize their performance over time based
on repetitive patterns in data. By continually learning from data patterns provided by
automated inspections, machine learning provides a roadmap for continuous
improvement.

Page | 4
Capitalizing on the data gained by automating the inspection process, brands, retailers,
factories, vendors and suppliers are all benefitting from the prescriptive insights machine
learning-based inspection platforms provide.
Automating inspections with machine learning is delivering the following benefits:
 Reducing risk, the potential for fraud, while improving the product and process
quality based on insights gained from machine learning is forcing inspection’s
inflection point.
 Matching the most experienced inspector for a given factory and product
inspection drastically increases accuracy and quality.
 Capture inspection data in real-time and know about potential and real quality
problems faster, using machine learning to spot anomalies in the data quickly.
 Improving track-and-trace accuracy, speed, and scale by being able to isolate the
specific root causes of assembly, materials and product quality problems early.
 Knowing why specific factories and products generated more Corrective
Action/Preventative Action (CAPA) than others and how fast they have been
closed in the past and why is now possible.
 Machine learning is making recommendations to inspectors on which defects to
look for first based on the data patterns obtained from previous inspections.
Image: How Machine Learning works

Page | 5
Inventory Optimization
Closely connected with supply chain optimization, machine learning can have a similar
impact on optimizing inventory. Holding costs (the cost of storing inventory) are massive,
usually hovering around 20-30% of the cost of a product. Even a modest reduction of
10% in holding costs can reduce per-unit costs by 2-3%. Holding unsold or undelivered
products means paying for storage space. This may not sound like a major problem, but
its effect on cash flow is immense.
Here, the role of machine learning is to calculate when it makes economic sense to hold
on to or sell inventory, or even increase/reduce production of inventory. This is done by
monitoring the supply chain elements mentioned above, as well as market prices, holding
costs, and production capacity. Carefully considering and balancing all of these elements
has traditionally been a human’s job. With the ever-increasing amount of data reflected
in each of these areas, however, humans are a poor choice for the task.
Therefore, the role of machine learning is an obvious one. By analyzing thousands or
even millions of bits of information to make decisions, these algorithms go far beyond
anything a human analyst is capable of. No surprise then that the results on overall
efficiency can be substantial.
Staring off, Machine Learning is an Artificial Intelligence based Project, typically a suite
of integrated applications that allows a company to collect, store, manage and interpret
data from many business activities through different algorithms, including Product
planning, Manufacturing or service delivery, Marketing and sales, Inventory management,
Shipping and payment.

Page | 6
Below is a detailed budget as to how the Machine Learning systems are to be
implemented and scheduling of costs over the span of 5 years.
Year 1
Server and Interface Set-Up $8,000
ML Development $8,000
Slippage $12,000
Licensing Fees $40,000
Infrastructure/Hardware $10,000
Total for Year 1 $78,000
Year 2
Maintenance Cost $30,000
Employees Recruitment/Training $40,000
Year 3
AI Upgradation $4,000
Year 4
Year 5
Total Cost required for each AI based Machine Learning system = $78,000 + $70,000
+ 34,000 + $34,000 + $34,000 = $ 2, 50,000 per Garments
 Note that the values are approximate and are used as estimates.

Page | 7
 Implementation Cost deals with how to estimate the cost, and how to avoid paying too
much. It includes analyzing and documenting specific business requirements, scope
and project plan. It also includes Installation, Configuration, Integration and
Customization.
 Infrastructure costs include purchase or upgrade of workstations.
 Maintenance cost is omitted from the first year of the plan and this has been taken
into account in all the other years. Because maintenance of the ML is extremely vital
since if the AI fails to work properly, the entire company will be in a standstill.
 The skilled employees of the company have to learn to operate the AI. As a result, a
god amount of money is also allocated for their utmost training so that they face no
difficulty in running the system.
 AI Upgradation comprises of the cost required in controlling the entire project to make
sure it stays on track and on budget.
Scheduling of costs:
Cost Year 1 Year 2 Year 3 Year 4 Year 5
Implementation Cost
Licensing Fee
Infrastructure
Employee Training
Maintenance Cost
AI Upgradation
How can ML address the internal challenges in the supply chain listed earlier?
 In the manufacturing phase, ML system helps capture and continuously monitor all
the actual data relating to quantities and costs, which will then be compared to the
budgeted/standard costs to determine inefficient production processes, managers and
workers, therefore providing good clarity into manufacturing and production
performance and quality, enabling on schedule production and sustaining highest
quality output.

Page | 8
 It aids identifying issues during the early phases which help to ensure efficient use of
production and employee assets.
 Reduces order cycle time
 Integration of business processes across departments onto a single information
system leads to low inventory, reduced operating cost, expedites delivery processes
and enables smoother work flow. It also eliminates shortage of resources and
ambiguity among departments.
 Allow faster decision making through real time AI.
 With access to all relevant data, partners can often answer questions themselves,
eliminating the need of other employees and conserving time.
 With the common database and integrated module, constant communication with the
buyers and suppliers can be maintained leading to an integrated and efficient supply
chain system for the organization.
 Enables having to spend less time for information procurement lead to better and
efficient utilization of time and resources.
 Eliminates the risk of duplicate work, and automates operational tasks, also leading
to efficient utilization of time.
The solution may not seem as lucrative as it sounds. There are limitations to the
implementation of Machine Learning system, mostly in terms of high maintenance costs
and the fact that most business owners take decisions based on profit margins rather
than future sustainability. Moreover, workers are not prone to changes and they rely on
various types of paper-based reports, unstructured decision making, changing roles, and
manual authorization procedures. Proper implementation of AI would require frequent
trainings of employees and system updates which can be costly for a company. Despite
the limitations involved, we firmly believe sincere cooperation and dedication among
workers will lead to successful implementation of the AI in all of the garment companies
we plan to aim further encouraging the rest of the industry to implement the like, hoping
to develop professional supply chain management practices and a boost in the growth of
the apparel industry.

Page | 9
Other recommendations:
 DIAD (Delivery Information Acquisition Device) can be used for proper tracking of the
shipments and their current condition and location by both manufacturer and buyer.
It’s a device which can make the overall delivery process faster and transparent.
Drawback: Sophisticated device, guidelines and training required for proper utilization.
 Collaboration of information using QR which focuses on POS to make supply chain
demand driven.
 Using RFID technology in order to maintain the real time tracing and visibility of the
garment products.
Total Words Counting: 1960

Page | 10
REFERENCE
“6 Ways Machine Learning Is Revolutionizing Manufacturing In 2019”
https://rapidminer.com/blog/6-ways-machine-learning-revolutionizing-manufacturing/
"Bangladesh's Ready Made Garments Landscape: The Challenge of Growth."
Http://www.mckinsey.de/sites/mck_files/files/2011_McKinsey_Bangladesh.pdf. Web. 14
May 2015.
"BEST-PRACTICE CASE STUDIES - UPS for Parcel Shipment and Tracking."
http://www.mobileinfo.com/case_study/ups.htm.Web. 15 May 2015.
Hasan, F. (2018, February 1). Are we ready for Industry 4.0? Retrieved May 1, 2019, from
Dhaka Tribune: https://www.dhakatribune.com/business/2018/02/01/ready-industry-4-0
“How Machine Learning Is Revolutionizing Inspections in Apparel Manufacturing”
https://www.inspectorio.com/blog/how-machine-learning-is-revolutionizing-inspections-
in-apparel-manufacturing/”
“How to Price an AI Project.” https://towardsdatascience.com/how-to-price-an-ai-project-
f7270cb630a4”
"Supply Chain Management on Apparel Order Process: A Case Study in Bangladesh
Garment Industry." http://www.dissertation.lib-ebook.com/d-economy/3063226-3-supply-
chain-management-apparel-order-process-case-study-bang.php.Web.15 May 2015.

Entering the 4th Industrial Revolution: Supply Chain Efficiency in RMG by using AI based Machine Learning Technology

More Related Content

What's hot (20)

Similar to Entering the 4th Industrial Revolution: Supply Chain Efficiency in RMG by using AI based Machine Learning Technology (20)

More from Akib Hasan Srabon (13)

Recently uploaded (20)

Entering the 4th Industrial Revolution: Supply Chain Efficiency in RMG by using AI based Machine Learning Technology