See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/374404004

Artificial Intelligence in Supply Chain Optimization

Conference Paper · June 2023

DOI: 10.1109/ICICAT57735.2023.10263631

CITATIONS READS
8 780

4 authors, including:

Anup Lal Yadav

Maharishi Markandeshwar (Deemed To Be University), Mullana Ambala, Haryana
67 PUBLICATIONS 158 CITATIONS

SEE PROFILE

All content following this page was uploaded by Anup Lal Yadav on 04 October 2023.

The user has requested enhancement of the downloaded file.

 ARTIFICIAL INTELLIGENCE IN SUPPLY
CHAIN OPTIMIZATION
Nisha Kumari Deepika Chaudhary Harmanjot Kaur Anup Lal Yadav
Department of Department of Department of Assistant Professor
Computer Science and Computer Science and Computer Science and Department of
2023 International Conference on IoT, Communication and Automation Technology (ICICAT) | 979-8-3503-0282-0/23/$31.00 ©2023 IEEE | DOI: 10.1109/ICICAT57735.2023.10263631

Engineering, Engineering, Engineering, Computer Science and

Chandigarh University Chandigarh University Chandigarh University Engineering,
Gharuan-140413 Gharuan-140413 Gharuan-140413 Chandigarh University
[email protected] [email protected] [email protected] Gharuan-140413
[email protected]
,

Abstract - With the advent of artificial intelligence (AI), better adapt and mitigate these issues [4]. However,
organizations can harness the power of machine learning technology and innovation are at the heart of supply chain
algorithms and big data analytics to gain greater insight into success. Therefore, the focus of the supply chain revolution
their supply chain processes and make more informed should be on optimizing and providing innovative supply
decisions. AI has proven to be a powerful tool for optimizing
chain solutions [5].
supply chain operations by automating processes, forecasting
demand and identifying areas for improvement. [1] Our
analysis of the main research poles reveals that learning
methods are slowly gaining ground, with observational and
interactive methods providing a new area of research. Finally,
we provide a roadmap for future research on AI applications in
SCM. Our analysis highlights the importance of behavioral
considerations in future research.
Keywords: Artificial Intelligence, Analysis Model,
Optimization.

I. INTRODUCTION

Supply chain management is an integral part of modern

business operations encompassing all activities associated
with sourcing, manufacturing, transporting, and delivering
goods and services to customers. Effective supply chain
management is crucial for organizations to maintain Fig.1 Modernization Of Supply Chain With AI
competitiveness, achieve cost efficiency, and meet
customer demand. Nonetheless, managing a dynamic and Advanced analytics and artificial intelligence are being utilized
intricate supply chain entails several challenges, such as by leading manufacturers to extract insights from distributed
intricate data management, supply chain disruptions, and datasets and develop solutions. This involves intelligent,
fluctuating demand. These issues can escalate into more secure, and transparent management of information related to
significant problems if not addressed appropriately [2]. processes, changes, and movements within the supply chain.
Therefore, it is vital to have improved visibility across the Blockchain technology plays a critical role in achieving
entire supply chain to help organizations respond promptly sustainability by enabling rational decision-making through
to changes in demand or supply, thereby reducing the risk artificial intelligence [6]. With the implementation of
of stockouts or overstocking [3]. To ensure future blockchain, there is a significant reduction in the need for
sustainability, supply networks must become more agile, authentication, leading to increased efficiency in supply chain
responsive, and traceable in response to disruptions. management, functional coordination, and collaboration
However, disruptions from diverse sources have prompted between businesses. Enterprises can encrypt transaction
companies to reassess their supply chain policies. Besides, records and store them in shared databases that are digitally
unforeseen supply chain disruptions are leading to transparent, securely protected from manipulation, alteration,
significant shifts in customer expectations and and erasure by interrelated technologies and trends.
unpredictable changes in demand and inventory costs. In
contrast, proactive data-driven supply chain ecosystems can

Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY SILCHAR. Downloaded on October 04,2023 at 01:59:44 UTC from IEEE Xplore. Restrictions apply.
 APPLICATIONS IN SUPPLY CHAIN Production and inventory management:
MANAGEMENT Production and inventory management encompass various
The subfields of artificial intelligence and their applications activities ranging from planning to delivery of finished
in the supply chain are explored in this overview. products. Efficient scheduling of resources and operations is a
crucial aspect of production management aimed at reducing
Demand forecasting and planning: production costs while meeting service objectives. Artificial
Demand forecasting and planning are critical functions in Intelligence (AI) techniques have been extensively used at all
various business areas such as inventory management, new levels of production management to optimize it [17]. For
product development, and marketing activities. instance, ANN has proven to be effective in hierarchical SC
scheduling, determining optimal batch sizes, and inventory
Key Management Issues: levels for demand and production planning. Rough set theory
However, traditional forecasting techniques such as moving has been used to reduce the bullwhip effect and optimize
averages, time series, and Box-Jenkins methods rely on past inventory management, while ML is employed in order
consumption to predict future demand, which limits their management, inventory management, and production
accuracy for new products or services without a history [7]. management to identify factors affecting inventory, optimize
Researchers are now turning to artificial intelligence reorder points and safety stocks, and estimate delivery times
algorithms to combine with traditional techniques to [18]. AI also helps organizations optimize logistics operations
improve demand planning accuracy, especially in the era of by selecting the most efficient routes and carriers to reduce
rapidly changing data [8]. transportation costs. The analysis of transportation data,
weather, and traffic information using AI algorithms enables
Artificial neural networks have been widely used for organizations to determine the most cost-effective and timely
forecasting and demand planning, with studies evaluating modes of transportation. In summary, AI-based supply chain
their performance in various industries such as oil management has revolutionized the optimization of operations
production, power generation by photovoltaic power plants, in organizations by minimizing information flow barriers and
and future prediction [9]. Machine learning is also resolving complex issues in real-time.
commonly used in demand forecasting and planning to
solve various supply chain problems, including predicting II. Literature Review
biased demand information [10], [11] managing the
bullwhip effect, and generating accurate predictions from
social media data. Additionally, ML algorithms are This literature review delves into the current state of research
extensively used in supply chain planning [12]. on using artificial intelligence (AI) in power chain operations
for optimization purposes, examining its benefits, challenges,
Procurement: and potential future directions [19]. AI, along with advanced
Procurement involves sourcing and acquiring raw materials, analytics, has the ability to analyze vast amounts of data from
supplies, services, and other items needed for production or different aspects of operations, including supply chains, to
sale, which requires companies to make decisions about enhance visibility. By implementing AI technologies, supply
manufacturing or outsourcing based on various factors such chain management can optimize inventory and reduce the
as production capacity, core competencies, and demand burden of decision-making, while maintaining high levels of
from suppliers [13]. Such decisions are often complex and accuracy, reliability, and authenticity [20]. Additionally,
challenging due to budget constraints, sourcing strategies, [21]AI systems can coordinate work sharing, negotiate with
and supplier relationships [14]. To address these challenges, suppliers, and complete tasks quickly and safely. However, to
researchers are exploring the use of artificial intelligence ensure security, an innovative supply chain access control
techniques to enhance procurement and supplier system is required to authenticate individuals, provide secure
management. access, and enforce access restrictions. AI, as a prominent
factor in the Fourth Industrial Revolution (4IR), [22] is capable
For instance, some studies have used artificial neural of outperforming human abilities in complex tasks, enabling
networks to improve supplier selection by enhancing organizations to quickly adjust plans and respond to changing
decision accuracy and update efficiency [15]. situations with flexibility, even in disrupted environments [23].
Other approaches such as Bayesian learning and risk
dependence modeling have been used to assess supplier A. Benefits of AI in Supply Chain Management
reliability and sustainability performance, respectively. AI has demonstrated several benefits for supply chain
Overall, these techniques aim to improve procurement management. These benefits include:
processes and help companies make better-informed
decisions in a complex and dynamic supply chain  Optimizing inventory management by reducing
environment. storage time and ensuring materials are moved
quickly from origin to destination [24].

Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY SILCHAR. Downloaded on October 04,2023 at 01:59:44 UTC from IEEE Xplore. Restrictions apply.
  Tracking shipments by providing real-time updates on
the location of goods at every point of contact.
 Improving order management by automating the process
of ordering what is needed when it is needed.
 Reducing costs by leveraging technology to find the
most efficient routes and using shorter distances to
deliver goods.
 Facilitating data analysis to make informed decisions in
real-time [25]. AI algorithms can analyze data faster than
humans, allowing for quick decision-making.
 Reducing manual labor by automating tasks such as
loading and unloading goods using AI-powered vehicles.

B. While AI has numerous benefits in supply chain

management, there are also some challenges that
need to be considered [26]:

• System complexity: AI systems usually require

extensive bandwidth and specialized hardware to access, Fig.2 Supply chain working
which can be costly and require a significant initial
investment for supply chain partners. • Data Collection: The initial step in any AI-driven supply
chain optimization project is to collect pertinent data. This
• Scalability factor: While AI and cloud systems are data may include information on customer demand,
scalable, the level of initial users/systems needed to inventory levels, production capacity, shipping routes, and
make a bigger impact and achieve efficiency must be other related factors. Data can be collected from various
discussed in depth with supply chain partners and their sources, including internal company systems, external
AI service providers. databases, and third-party providers.

• Training cost: As with any new technological solution, • Data Cleaning and Pre-processing: Once data is
training is a significant investment in terms of time and collected, it is necessary to clean and pre-process it to
money. This can affect business efficiency, and supply ensure accuracy and consistency. This involves removing
chain partners must work with AI providers to create an duplicate or irrelevant data, correcting errors, and
effective and affordable training solution during the standardizing formats. The data may also need to be
integration phase. converted into a format that AI algorithms can utilize, such
as numeric data.
• Associated operating costs: AI-operated machines have
individual processors that require maintenance and • Model Selection: The subsequent step is to select an
replacement from time to time, which can be expensive appropriate AI model or algorithm for a particular supply
and increase utility bills, thereby impacting overhead chain optimization task. This may involve the utilization of
costs. Manufacturers would have to consider these costs machine learning algorithms such as regression analysis,
to keep the machines running. decision trees, or neural networks. The selection of the
model depends on the specific project goals and available
data.
III. METHODOLOGY
• Model Training: After selecting a model, it must be
The methodology employed in utilizing artificial trained using historical data. This entails feeding the model
intelligence to optimize supply chain management can vary with data and adjusting its parameters to optimize
depending on the specific project goals and application. performance. The training process can take some time,
However, certain depending on the model's complexity and the amount of
common approaches and techniques are often used in AI- data available.
driven supply chain optimization.
• Model Validation: Once the model is trained, it must be
validated against test data. This includes evaluating the
model's performance in predicting outcomes based on new
data. The validation process is crucial to ensuring the
model's accuracy and reliability.

Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY SILCHAR. Downloaded on October 04,2023 at 01:59:44 UTC from IEEE Xplore. Restrictions apply.
 • Deployment: Once the model is validated, it can be • Increased agility: Artificial intelligence can provide real-
deployed in a production environment. This may time data and insights, enabling organizations to respond
involve quickly and effectively to changes in the market or supply
integrating the model into existing supply chain chain. This can enhance organizational agility and improve
management systems or developing new systems to responses to new opportunities and challenges.
support the model's predictions. The deployment
process may also include monitoring model V. CONCLUSION
performance and making adjustments as needed.
In conclusion, the utilization of artificial intelligence (AI) in
• Maintenance and Updates: Finally, the AI model
supply chain management can offer several advantages for
needs to be maintained and updated continuously to
businesses. Through the analysis of historical data, AI
ensure that it continues to provide accurate predictions.
algorithms can forecast demand and adjust inventory levels,
This may include monitoring model performance,
reducing the risk of stockouts and overstocking. Furthermore,
collecting new data, and updating model parameters as
AI can optimize logistics and transportation networks by
needed.
analyzing traffic patterns and other relevant factors, leading to
lower shipping costs and improved delivery times.
IV. OUTCOMES Additionally, AI can aid in supplier selection and management
by identifying the best suppliers and improving contract
Using artificial intelligence (AI) to optimize supply chain negotiations, resulting in better quality and lower costs. Lastly,
management can yield several benefits for organizations. AI can help organizations improve their environmental impact
Here are some common outcomes of applying AI in supply by analyzing data on energy consumption and carbon
chain management: emissions. As AI technology advances, we can expect to see
even more benefits in the future. However, the widespread
• Enhanced efficiency: AI can optimize supply chain application of AI in supply chain management depends on the
processes and reduce lead times, enabling development of advanced computer chip technology to track
organizations to respond promptly to customer transportation, which requires further research and analysis of
demands. This can result in faster delivery times, data. Additionally, automation of consumer interactions is a
reduced costs, and increased customer satisfaction. new area that has potential in marketing.

• Improved demand planning: AI can accurately VI. FUTURE SCOPE

forecast demand by analyzing historical data and other
relevant factors. This can assist organizations in
The future scope of optimizing supply chain management using
avoiding overstocking or understocking, minimizing
artificial intelligence is vast and offers numerous potential
waste and improving profitability.
applications across various industries and fields. Some of the
most promising areas of development include predictive
maintenance, autonomous transportation, risk management,
• Enhanced visibility: Artificial intelligence can provide
sustainability, and collaboration.
real-time insights into inventory levels, production
status, and shipping routes, enabling organizations to
One area of development is predictive maintenance, where AI-
make better decisions and respond quickly to
powered tools can predict equipment failures before they
unexpected disruptions in the supply chain.
occur, allowing for preventive maintenance to reduce
downtime and repair costs. This can be especially beneficial in
• Cost savings: AI can identify opportunities for savings
industries where equipment failure can have significant
by optimizing production schedules, reducing
transportation costs, and minimizing waste. This can financial and safety consequences, such as manufacturing and
transportation.
lead to significant cost savings over time.
Another area of development is autonomous transportation,
where AI-powered vehicles such as drones and autonomous
• Improved collaboration: AI can facilitate better
vehicles can transport goods across the supply chain. This can
collaboration between organizations, suppliers, and
reduce labor costs and shorten delivery times. AI tools can
customers by sharing information and coordinating
enable these vehicles to navigate obstacles and make real-time
activities. This can enhance communication and foster
decisions based on changing conditions, improving efficiency
stronger relationships, leading to better outcomes for
and safety.
all parties involved.

Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY SILCHAR. Downloaded on October 04,2023 at 01:59:44 UTC from IEEE Xplore. Restrictions apply.
 Risk management is also an area where AI can provide [7] K. N. Amirkolaii, A. Baboli, M. K. Shahzad, and R.
significant benefits. AI tools can help organizations Tonadre, “Demand Forecasting for Irregular Demands in
identifypotential risks in the supply chain, such as weather Business Aircraft Spare Parts Supply Chains by using
events or political instability, and develop contingency plans Artificial
to mitigate those risks. This can help organizations better Intelligence (AI),” IFACPapersOnLine, vol. 50, no. 1, pp. 15221–
15226, Jul. 2017.
prepare for unforeseen events and minimize the impact of
[8] A. Gligor, C. D. Dumitru, and H. S. Grif, “Artificial intelligence
disruptions. solution for managing a photovoltaic energy production unit,”
Procedia Manuf., vol. 22, pp. 626–633, Jan. 2018.
Sustainability is another area where AI can help optimize [9] L. B. Sheremetov, A. González-Sánchez, I. LópezYáñez, and A.
supply chain management. AI-based tools can help V. Ponomarev, “Time Series Forecasting: Applications to the
organizations reduce environmental impact by optimizing Upstream Oil and Gas Supply Chain,” IFAC Proc. Vol., vol. 46,
shipping routes, reducing waste, and identifying sustainable no. 9, pp. 957–962, Jan. 2013.
suppliers. This can not only help the environment but also [10] P. K. Bala, “Improving inventory performance with clustering
improve the organization's reputation among based demand forecasts,” J. Model. Manag., vol. 7, no. 1, pp. 23–
environmentally conscious customers. 37, Jan. 2012.
[11] N. A. Mobarakeh, M. K. Shahzad, A. Baboli, and R. Tonadre,
Finally, AI-based tools can enable collaboration between “Improved Forecasts for uncertain and unpredictable Spare Parts
Demand in Business Aircraft’s with Bootstrap Method,”
various stakeholders in the supply chain, such as suppliers,
IFACPapersOnLine, vol. 50
logistics providers, and customers. By improving [12] H. Wenzel, D. Smit, and S. Sardesai, “A Literature Review on
communication and coordination, organizations can reduce Machine Learning in Supply Chain Management Supply Chain
inefficiencies and respond more quickly to changing Management,” in Proceedings of the Hamburg International
demands and unexpected disruptions. Conference of Logistics (HICL) - 27, 2019, no. September, p. 30.
[13] H. Bousqaoui, S. Achchab, and K. Tikito, “Machine learning
Overall, the future scope of optimizing supply chain applications in supply chains: Long shortterm memory for demand
management using artificial intelligence is significant, and forecasting,” in Lecture Notes in Networks and Systems, 2019.
its applications are likely to continue to expand across [14] M. T. Hoppe, “Machine Learning in Supply Chain
industries and fields. The potential benefits of AI-based Management,” no. August, 2019
supply chain management tools, such as improved [15] K. Zhao and X. Yu, “A case based reasoning approach on
efficiency, reduced costs, and increased sustainability, make supplier selection in petroleum enterprises,” Expert Syst. Appl.,
2011
it an exciting area of development for businesses and
[16] C. Jiang and Z. Sheng, “Case-based reinforcement learning
organizations seeking to remain competitive in a rapidly
for dynamic inventory control in a multiagent supply-chain
changing global marketplace. system,” Expert Syst. Appl., 2009
[17] M. D. Garvey, S. Carnovale, and S. Yeniyurt, “An analytical
VII. REFERENCES framework for supply network risk propagation: A Bayesian
network approach,” Eur. J. Oper. Res., 2015.
[1] Kusi-Sarpong, S., & Huang, X. (2020). Artificial intelligence [18] M. Mehdizadeh, “Integrating ABC analysis and rough set theory
applications in supply chain management: A review and future to control the inventories of distributor in the supply chain of auto
research directions. Expert Systems with Applications, 153, spare parts,” Comput. Ind. Eng., 2020.
113406. [19] Helo, P.; Hao, Y. Artificial intelligence in operations
[2] D. Nguyen and W. Bernard, “Improving the learning speed of management and supply chain management: An exploratory case
2-layer neural networks by Choosing initial values of the study. Prod. Plan. Control 2021, 33, 1573–1590.
adaptive weights,” IJCNN International Joint Conference on [20] Kashem, M.A.; Shamsuddoha, M.; Nasir, T.; Chowdhury, A.A.
Neural Networks, vol. 1-3, pp. C21–C26, 1990. The Role of Artificial Intelligence and Blockchain Technologies
[3] T. El-Ghazali, Meta Heuristics-From Design to in Sustainable Tourism in the Middle East. In Worldwide
Implementation, John Wiley and Sons, Hoboken, New Jersey, Hospitality and Tourism Themes; Emerald Publishing Limited:
2009. Bingley, UK, 2022; pp. 1755–4225. ISSN 1755-4217.
[4] S. K. S. Fan and E. Zahara, “A hybrid simplex search and [21] Pimenidis, E.; Patsavellas, J.; Tonkin, M. Blockchain and
particle swarm optimization for unconstrained optimization,” Artificial Intelligence Managing a Secure and Sustainable Supply
European Journal of Operational Research, vol. 181, no. 2, pp. Chain. In Cybersecurity, Privacy and Freedom Protection in the
527–548, 2007. Connected World; Springer: Berlin/Heidelberg, Germany, 2021;
[5] X. Huang, B. L. Tan, and X. Ding, “An exploratory survey of pp. 367–377.
green supply chain management in Chinese manufacturing [22] Gohil, D.; Thakker, S.V. Blockchain-integrated technologies for
small and medium-sized enterprises: Pressures and drivers,” solving supply chain challenges. Mod. Supply Chain Res.
Computational Intelligence and Neuroscience 9 Journal of Appl. 2021, 3, 78–97.
Manufacturing Technology Management, vol. 26, no. 1, 2015. [23] Zhang, Z.; Song, X.; Liu, L.; Yin, J.; Wang, Y.; Lan, D. Recent
[6] M. Clerc, Particle Swarm Optimization, U.K.Wiley ISTE, advances in blockchain and artificial intelligence integration:
London, 1st edition, 2006. Feasibility analysis, research issues, applications, challenges, and
future work. Secur. Commun. Netw. 2021, 2021, 9991535.

Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY SILCHAR. Downloaded on October 04,2023 at 01:59:44 UTC from IEEE Xplore. Restrictions apply.
 [24] Pham, D.T, A.A. Afify, Artificial Intelligence techniques
and their applications in manufacturing, April 2005,
Proceedings of the Institution of Mechanical Engineers Part B
Journal of Engineering Manufacture 219(5): 395–412.
[25] Glass, K. and Colbaugh, R., 2013. Improving supply chain
security using big data. In: K. Glass, ed. 2013. IEEE
International Conference on Intelligence and Security
Informatics (ISI), 2013. Piscataway, NJ, pp. 254–259.
[26] Yan, W., He, J. and Trappey, A., 2019. Risk-aware supply
chain intelligence: AIenabled supply chain and logistics
management considering risk mitigation. Advanced
Engineering Informatics, 42, p.100976.

AUTHORS PROFILE

Nisha Kumari is a student of Chandigarh

University, Punjab pursuing Bachelor of
Engineering Computer Science and in
Engineering 2019-2023. She had completed her
AISSCE from SHMD School, Jaigaon, West
Bengal in the Year 2019. Her research interest is
in computer application & trending technologies.
Apart from these she is also interested in sports, art and calligraphy
and got certified at national level for handwriting.

Deepika Choudhary is a student of Chandigarh

University, Punjab pursuing Bachelor of
Engineering in Computer Science and Engineering
2019-2023. She had completed her AISSCE from
arsh Shiksha Niketan School, Jammu in 2019. She
has worked in projects using API and makes many
software clones. Apart from this she is interested
in art and reading.

Harmanjot Kaur is a student of Chandigarh

University, Punjab pursuing Bachelor of
Engineering Computer Science and in
Engineering 2019-2023. She had completed her
AISSCE from Sidhu Memorial Public School
Sunam-Mansa Road, Punjab in the Year 2019.
Her research interest is in computer application &
trending technologies. Apart from these she is
also interested in sports, art and got certified at state level in many
arts competitions.

Anup Lal Yadav is persuing his PHD from

MMU Mullana, Ambala, India. He has 13 years
of teaching experience. Currently, he is working
as assistant professor at Chandigarh University,
Gharuan, Mohali, Punjab, India. He has
published more than 35 research papers in
reputed journals. He is a member at IEEE, CSTA
and AIENG. His research interest areas are IOT, AI, Cloud
computing, Machine Learning.

Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY SILCHAR. Downloaded on October 04,2023 at 01:59:44 UTC from IEEE Xplore. Restrictions apply.

View publication stats