Artificial Intelligence in Supply Chain Management

Artificial Intelligence in Supply Chain

Management

SCM 463 - Global Supply Chain

Management

Arizona State University-Tempe

December 14, 2022

Introduction

Background

Industry 4.0, which refers to the fourth

industrial revolution and also known as

4ILR, is a digital transformation in the

field of manufacturing, production, and

processing that emphasizes the need for

the use of artificial intelligence (AI) when

making decisions and undertaking

specific actions. Information has become

a critical factor that defines the success or

failure of a given business (Yuanita, 2019,

p. 15). As such, firms are under pressure

to find the best ways of processing large

data within the shortest period possible

and use it to make critical decisions.

Traditionally, humans would be expected

to process the data using simple

machines and then make relevant

decisions. However, that is changing with

the emergence of concepts of Big Data

and Machine Learning (ML). It is

becoming increasingly possible to rely on

machines to collect and process data, and

then use it to make critical decisions and

predictions more accurately than a

human would (Tirkolaee, Sadeghi,

Mooseloo, Vandchali, & Aeini, 2021, p.

13). As such, AI has gained popularity in

the field of supply chain management

(SCM). In this study, the researcher

focuses on the benefits, opportunities,

and challenges of embracing ML

technologies in logistics to help SMEs

boost their performance.

Research Problem

Micro, Small & Medium Enterprises

(MSMEs) play a critical role in emerging

and developing economies in terms of

the creation of employment and the

growth of gross domestic products (GDP).

However, they face unfair competition

from large multinational corporations

that have the financial power,

experience, and the right connections to

facilitate their growth and embrace

emerging technologies. Adopting

industry 4.0 has been a major challenge

for these small and medium enterprises

because of the financial implications and

expertise needed. Despite the challenge,

Singh, Wiktorsson, and Hauge (2021)

explain that industry 4.0 offers a unique

opportunity for these firms to redefine

their operations and compete favorably

with the more established rivals (p. 67).

Embracing AI in SCM makes it possible to

reduce expenses on human resources,

improve efficiency in operations, and

enhance the quality of products that they

deliver.

Research Objectives

This study aims to establish a practical

and clear of the benefits, opportunities,

and challenges of adopting machine

learning technologies in logistics to help

SMEs boost their performance. To

achieve this primary goal, the researcher

will focus on realizing the following

objectives:

1. To identify ML techniques

frequently used in logistics in SMEs;

2. To identify the benefits of ML in

Logistics in SMEs;

3. To identify the opportunities of

implementing ML techniques in

logistics in SMEs;

4. To identify the challenges of

implementing ML techniques in

logistics in SMEs;

5. To identify the relationship

between ML adoption in logistics and

MSMEs performance.

Research Questions

Artificial intelligence is a field that has

continued to attract the attention of

many scholars over the years. It is

essential to specifically define the specific

issues that need to be investigated in a

given study. In this research, the

following are the specific questions that

the researcher seeks to answer:

1. What are ML techniques

frequently used in logistics in SMEs

and what are the benefits and

opportunities of adopting them?

2. What are the challenges of

adopting ML techniques in logistics in

SMEs?

3. What is the relationship between

ML adoption in logistics and MSMEs

performance?

Overview of the Current State of

Research and Methodology

Artificial intelligence and its relevance in

supply chain management is a field of

study that has attracted massive

attention from scholars over the years.

According to Singh et al. (2021), scholars

have been keen to understand how AI

can help MSMEs to acquire materials

they need for production more efficiently

and cost-effectively (p. 69). However, a

preliminary review of existing studies

shows that this is a relatively new field as

AI continues to evolve. In this project, the

researcher will rely on secondary data

sources to help achieve the aim and

objectives identified above.

Evolution of Artificial Intelligence

Artificial intelligence and machine

learning have become increasingly

relevant in the modern business

environment. It is necessary to discuss

the evolution of this technology and how

it gained relevance in different fields

(Rabayah, 2013, p. 8). The origin of AI can

be traced back to Europe in the early

20th century, especially in the United

Kingdom and Germany. Mathematicians,

scientists, and philosophers have

conceptualized the idea of having

intelligent machines capable of solving

human problems in the 1950s. Several

scientists were able to develop a logical

framework in 1950 referred to as

Computing Machinery and Intelligence,

which would later redefine the current

use of AI in supply chain management

(Pervaiz, 2020, p. 26). He explained how

machines can be trained to be intelligent

and ways in which their intelligence can

be tested.

Stages of AI Growth

The pioneers in this industry faced

numerous challenges that slowed the

growth of this industry. One of the

biggest challenges was the accessibility of

the computers at this time. Sendler

(2017) explains that leasing a computer in

the 1early periods of its development

was an expensive undertaking, which

meant that only prestigious and

financially empowered institutions could

afford it (p. 87). He describes computers

at that time as being primitive, only

capable of executing commands but not

storing them. Many institutions did not

see the worth of investing in this

technology, which means that these

pioneers did not get the relevant support

they needed to facilitate the growth of

the industry. However, they were able to

overcome these challenges and

facilitated its gradual growth.

It became evident that the growth of AI

needed experts to share ideas to help in

improving its capabilities. The Dartmouth

Summer Research Project on Artificial

Intelligence (DSRPAI) of 1956, which was

hosted by Marvin Minsky and John

McCarthy was the first of its kind and

catalyzed its growth. It brought together

experts in AI and financiers from all over

the world. The technology started

developing as machines could not only

execute but also store data and

remember actions taken.

The Fifth Generation Computer Project

(FGCP), which started in 1980 and lasted

till 1990, saw major improvements in the

technology as computer programs

became enhanced. Personal computers

had also become common, significantly

reducing the cost of having computers

(Matt, Modrák, & Zsifkovits, 2020, p.

310). By the late 1990s, machine learning

had started gaining massive popularity

around the world. The success of Deep

Blue and Alpha Go in the gaming field

demonstrated that AI had achieved a

level of intelligence that matched or even

surpassed that of humans. During the

same period, the concept of big data also

started gaining rapid popularity. Figure

2.1 below summarizes the evolution path

of artificial intelligence.

Figure 2.1: Evolution of AI

The growth of AI in the late 20th century

and the 21st century can be summarized

in three stages, as shown in figure 2.2

below. The first and critical stage was

machine learning. Large technology

companies such as Apple Inc., Microsoft,

and Amazon.com created Siri, Cortana,

and Alexa respectively. Small companies

may not have such sophisticated tools

but they have a role to play in the general

development of the industry

(Madanchian, Hussein, Noordin, &

Taherdoost, 2015, p. 78). These were

tools capable of communicating with

humans and helping people to execute

specific commands such as conducting

online research and storing specific

information, helping eliminate many

challenges in the supply chain (Linh,

Kumar, & Ruan, 2019, para. 117). These

tools were also designed to become

supper-collectors of data to help these

firms to understand the emerging needs

and expectations of customers. They

were designed to learn based on

information that they gathered

independently.

The success of machine learning led to

the second stage of machine intelligence.

These machines were becoming

increasingly independent as they could

collect data and execute specific actions

without any assistance from human

beings. Earley (2014) explains that the

current technology is at the level of

machine intelligence (p. 59). Once the AI

is trained on basic tasks, they can

facilitate their further training as long as

they can access accurate and reliable

data. These machines are also capable of

taking specific actions without any

human input. The predictions and

decisions that they make are sometimes

superior to that of humans. This is so

because they wholly rely on data instead

of personal biases and cultural

misconceptions, which enhances their

competitiveness (Jovanovski, Seykova,

Boshnyaku, & Fischer, 2019, p. 250). In

the supply chain sector, this technology is

proving to be crucial in supporting the

decisions that a firm makes, based on

data.

The next stage of AI, known as machine

consciousness, is still futuristic. Currently,

AI primarily bases its decisions on data,

focusing on lowering costs while

maximizing profitability (Devang,

Chintan, Gunjan, & Krupa, 2019, p. 31).

However, humanity goes beyond

amassing wealth as these machines do.

Compassion and the need to help those

in distress make humans unique in their

decision-making process. The fact that

these machines lack consciousness may

sometimes force managers to ignore the

predictions, decisions, and simulations of

AI. A firm may decide to invest in

corporate social responsibility (CSR) to

help the needy even if the plan

developed by AI was different. The

current progress in this field of

technology is to facilitate the

development of machine consciousness

(Giudice, Scuotto, Garcia-Perez, &

Petruzzelli, 2019, p. 310). AI should have

human feelings and act in the interest of

humanity.

Figure 2.2: Stages of AI development

Application of AI

AI creators started spreading the

application of AI from the gaming

industry, where it had achieved massive

success, to the business sector. It was

evident that AI could help firms in making

accurate decisions and predictions when

it is fed with the right data (Chung, 2021,

p. 55). One of the areas that it gained

massive popularity was in the

manufacturing sector. As market

competition became stiff, firms were

under pressure to find production

strategies that would lower the cost

while enhancing quality and productivity

as a way of improving profitability. AI

made it possible to develop simulations

of the best manufacturing strategies.

Robots also started taking active roles in

the manufacturing plants as a way of

lowering costs, reducing risks to humans

at these plants, and increasing

standardization and productivity. The

success of AI in the manufacturing sector

made business executives find ways of

applying it in other sectors.

Marketing became another major area

where AI technology gained popularity.

Firms needed to accurately predict

market demands, changing tastes and

preferences, and ways of meeting clients’

needs in the best way possible. AI

became a critical tool that facilitated

predictions and decision-making based

on data. Small and medium-sized

enterprises are under immense pressure

to find ways of improving their efficiency

(Alfoqahaa, 2018, p. 12). AI offers them a

perfect way of overcoming numerous

operational challenges. It can monitor

the money markets and help in making

accurate investment decisions. It also

became an essential financial planning

tool for small, medium, and large

corporations. Intelligent robots are also

currently used in supply chain

management, which is the primary focus

of this paper. Other areas where AI is

actively used include green

manufacturing and other environmental-

friendly initiatives. Figure 2.3 below

identifies areas where AI has become a

crucial tool that facilitates effective

operations.

Figure 2.3: Application of AI

Technologies, Security, People/Workers

and Society

The rapid growth of AI and its current

application in various fields has been

facilitated by the desire to achieve

efficiency and lower costs. In AI, volume,

velocity, and variety have been the

primary factors that have enhanced

growth. Data has become a critical factor

that defines the ability of a firm to

achieve success. However, firms find

themselves in situations where they have

to collect large amounts of data to help in

the decision-making process. Irrespective

of a firm’s size, it is essential to collect

relevant data that can help inform

decisions and predictions that a company

makes. AI makes it possible to collect

large amounts of data, process it based

on its relevance, and use it to conduct

simulations and predictions (Aarstad &

Saidl, 2019, p. 39). The capacity of the AI

to handle large volumes of data surpasses

that of humans in a significant way.

The speed or velocity with which data is

processed is another important factor

that has made AI popular in the business

world. When collecting data, it is

common for one to have access to large

amounts of information. Humans are

limited in terms of the amount of data

that they can process within a specific

period. The problem with such limitations

is that sometimes the information that is

ignored is the most crucial for the firm’s

growth (Kumar, 2019, para. 6). AI makes

it possible to process such large amounts

of data within a short period. It can

speedily select the most relevant data,

based on the issue that it has to address

and then process the data to help solve

the problem. Once processed, AI also

makes it easy to share information with

all the relevant stakeholders in real-time.

AI and ML enable a firm to have a variety

of options when deciding on various

issues. For instance, when a firm is

interested in selecting an appropriate

supplier, various factors such as price,

quality, and reliability have to be

considered. It is common to find a case

where the decision is solely based on

price or quality instead of the other

equally important variables. Stakeholders

may not feel comfortable when they are

not aware of the decision-making process

(Adixon, 2019, para. 9). AI takes into

consideration all the other important

factors through simulation. It then

provides various alternatives, in terms of

priority, that the firm can consider. It

evaluates both pros and cons in the

simulation before classifying the possible

actions that a firm can take to achieve the

intended goal.

This technology is also meant to make the

work of people, especially employees,

easier than it was in the past. Some of the

physically and mentally demanding tasks

can now be done using machines. Robots

have become effective in handling

physically demanding tasks. On the other

hand, sophisticated computers can now

handle complex data processing and help

in making accurate decisions. According

to Amblee (2018), AI is emerging as a tool

that will help protect modern society

from natural forces such as global

warming and climate change (para. 11).

Fed with the right data, they can facilitate

smart manufacturing and transportation

in ways that have negligible impact on the

environment. It can be possible to

reverse some of the damages that have

been made to the environment.

Scientists believe that AI is set to make

society a better place.

Artificial Intelligence in Supply Chain

Management

Artificial intelligence has become a

critical tool in supply chain management.

According to Anyoha (2017), news about

having driverless trucks was welcome

news to many trucking companies in the

United States (para. 7). Although the idea

is yet to be actualized, its development is

in advanced stages and there is a

consensus that most cars on the road will

not need drivers. For major trucking

companies around the world, such a

technology will have an immense impact

in cutting the cost of hiring numerous

drivers, paying their medical allowance,

and having to deal with workers’ unions.

AI goes beyond the promised creation of

driverless trucks (Kersten, Blecker, &

Ringle, 2019, p. 18). Numerous benefits

are already available for firms that have

chosen to use the new technology. It is

necessary to discuss how AI and ML have

gained relevance in SCM and how they

are applied to help cut operational costs,

improve efficiency, and enhance

profitability.

Data and Optimization across the Value

Chain

The concept of data optimization has

gained massive relevance in the recent

past as firms embrace big data. Data

optimization refers to a process that uses

sophisticated data management tools to

have access to, organize, and process

data from various sources at high speeds

and in a comprehensive manner (Wright

& Recht, p. 45). The processed data is

then used to make critical decisions

within an organization. The primary goal

of data optimization is to facilitate a

performance that is capable of meeting

customers’ expectations (Dash,

McMurtrey, Rebman, & Kar, 2019, p. 43).

It makes it possible to provide real-time

service to clients by understanding their

emerging needs.

In supply chain management, data

optimization has become essential in

ensuring that smart manufacturing is

achieved irrespective of the size of a firm.

Didonet and Díaz (2012, p. 101) observe

that successful companies have come to

appreciate the significance of big data

and its application in making predictions

and decisions. Smart manufacturing

requires effective communication and

coordination in the entire supply chain,

from the stage of raw material

acquisition to the stage when the product

is delivered to clients. As shown in figure

3.1 below, there are five stages that are

involved in smart manufacturing when AI

and ML are applied. The initial stage

involves collection of raw data. At this

stage, the focus is to gather relevant

information about different suppliers to

understand those that offer superior

value.

The second stage is the visualization and

integration of data. A firm starts to

develop a database that can help it in

making critical decisions. The third stage

of material-centric insight involves the

systematic use of the collected data to

assess the suppliers’ worth to the firm

(Belhadi, Mani, Kamble, Khan, & Verma,

2021, p. 4). Continued reliance on data

facilitates transformative analytics and

continuous improvement. At this stage, a

firm will be moving from traditional

approaches of manufacturing to one that

integrates the use of machines in making

predictions and decisions. Once the data

is available, a firm can trust computers to

process it and use it to accurately address

specific challenges that affect the

operations of a company. When these

four stages are completed successfully,

the company can move to the final stage

of having fully automated smart

manufacturing. At this stage, computers

will be responsible for providing

simulations about the best ways of

operation that can yield the best output

for the company.

Figure 3.1: Stages in smart manufacturing

AI and ML have made it possible for firms

to achieve fully automated smart

manufacturing. Traditionally, firms had to

make manufacturing decisions based on

human predictions (Toorajipour,

Sohrabpour, Nazarpour, Oghazi, Fischl,

2021, p. 506). The problem with such

predictions was that in most of the cases

it was based on outdated data and

personal biases that led to consistent

inaccuracies. It was always impossible to

produce the exact number of products

needed by clients. The products were

either in excess or less than what the

market needed. Smart manufacturing

helps in eliminating such challenges.

Real-time data makes it possible for a

firm to understand the current demand in

the market.

The information will then be used to

facilitate the delivery of the exact amount

of raw materials needed, which will then

be fully utilized to develop products. The

produce will then be delivered to the

market within the right time. Smart

manufacturing helps in eliminating

wastes caused by overproduction and

delays that result from underproduction.

It also helps in ensuring that the right

quality of products is made available to

customers at the right time. AI and ML

make it possible for smart manufacturing

to be realized in an organization.

ML Techniques Frequently Used in

Logistics in MSMEs

Machine learning has achieved

popularity in various business settings

because of its unique capabilities. For

small and medium enterprises, this

technology offers a wide range of

services that can be used to enhance

activities in the logistics sector. Figure 3.2

below identifies specific ML techniques

that can be used in logistics. Face

recognition has become a powerful tool

that firms use to facilitate the

identification of persons. In the logistics

sector, one of the challenges that SMEs

face is cases where unauthorized

individuals have access to goods on

transit for the primary purpose of stealing

(Chin, Hamid, Rasli, & Baharun, 2012, p.

615). ML has introduced a new

technology that is meant to enhance the

security of products while they are in

transit. Face recognition is a technology

that requires drivers and authorized

individuals to show their faces before

special in-built cameras on the trucks

before one can access the cargo section.

It means that unauthorized individuals

cannot have access to the cargo section

of these trucks. When one tries to make

a forced entry without using face-

recognition technology, information is

sent directly to the security agencies.

The technology can also be used to

ensure that drivers do not allow

unauthorized passengers into the cabin

because they can also be a security threat

to the employee and goods on. The tool

is also widely used in enhancing security

at warehouses transit (Singh, Kumar, &

Shankar, 2012, p. 174). Only those whose

faces the machine can recognize are

allowed into the warehouse. The

technology not only helps in eliminating

the possibility of unauthorized persons

having access to goods and raw materials

but also makes it possible to identify

employees who are stealing from the

firm. The technology records time an

employee entered the warehouse or the

cargo section of the truck, the time they

took before leaving, and what they

carried when leaving. The chances of

stealing from a firm are significantly

reduced in such instances, which means

that the cost of operation will be dropped

significantly.

Object detection is another ML technique

that is gaining relevance in logistics. Using

special sensors, the machine can detect

when an object is approaching and take

necessary action. The technology is

currently used to automate various

activities in warehouses (Lawson, 2021,

para. 7). A common application is

controlling the security lights. Instead of

security lights being on at all the sections

of the warehouse, the sensor can be

trained to detect when people are in the

store and it will automatically switch on

the lights. When people are not within

the warehouse, the sensor automatically

switches off the lights. The technology

helps in saving the cost of energy. Object

detectors can also be used to enhance

security.

When an object, which in most cases is a

human, is detected moving into a

warehouse, information will be relayed

to the security unit immediately. If the

object is considered suspicious,

immediate action can be taken to inspect

it and determine if indeed it is a security

threat. Smart driving technology is

making it possible for trucks to identify

objects on their paths. The technology,

although not fully developed, makes it

possible for a vehicle to make emergency

breaks when it is about to hit an object on

its path. Such technologies will

significantly increase safer working

conditions (Jacobs, 2020, para. 13). It will

significantly reduce losses of lives, trucks,

and goods on transit caused by these

accidents. The insurance premium that

MSMEs have to pay on such smart trucks

and goods on transit will be significantly

low. Object detection technology is

directly related to motion detection, both

of which work together to enhance the

capability of ML.

Advanced ML technology is currently

capable of emotion recognition. When a

driver, a captain, a pilot, or warehouse

employees are in their place of work, AI is

capable of detecting their emotions. The

ML can recognize when one laughs a lot

or specific words that they use in their

conversation to express their emotions.

Advanced AI has become critical in

making important supply chain

management decisions (Stinson, 2021,

para. 3). They can tell when one is

distressed or frustrated while they are on

their official duties. Mental stress

compromises the ability of one to reason.

Emotion recognition technology is meant

to eliminate such cases by identifying

employees who are emotionally

distressed while at work. When they are

captains, drivers, or pilots, measures can

be taken to ensure that they are not

allowed to be fully in control of the

transport vessel. If possible, such an

individual should be temporarily relieved

of their duties to get the relevant

assistance before they can resume work.

When it is not possible or the threat is

assessed to be negligible, they can be

allowed to work alongside a colleague

who can help them navigate the vessel

and to talk about the issue of concern.

ML has made it possible to convert text to

speech in a conversation. Texting while

driving is a major cause of road accident

in the United States. It is significantly

more dangerous than making a call

because one has to take their eyes off the

road to write a text (Pournader, Ghaderi,

Hassanzadegan, & Fahimnia, 2021, p. 2).

Such an action makes it easy for a driver

to lose focus on the road. ML has

introduced a technology where a text can

be converted easily into a sound and vice

versa. It means that a driver can

communicate with others easily without

having to constantly take their eyes off

the road. Their speech can then be

converted to a text in case the person on

the other end of the phone prefers

chatting. Another benefit of this

technology is that a driver, a captain, or a

pilot can make their reports easily

without having to type them. At every

stage, they can verbally report what has

taken place at a specific time. Their verbal

report is then transformed into texts,

which can then be documented for

further references. It eliminates the need

to develop a written report, which saves

them time.

Automatic speech recognition is another

aspect of ML that is proving to be crucial

in the logistics sector. Like face

recognition technology, speech

recognition is also gaining relevance as a

tool that can be used to identify

employees within a firm (Hellingrath &

Lechtenberg, 2019, p. 67). The AI has

developed a unique capacity that enables

it to identify a unique sound of an

individual, just like a fingerprint. Unlike a

signature that can be forged, one cannot

defeat a well-trained machine by faking

another person’s voice. It is also not

possible to steal the voice as one would

steal a password.

One would need to say their name or

utter a specific statement when they

want to have access to the warehouse or

a transport vessel. The voice recognition

tool will then determine if they are

authorized to have access to the store or

the vessel. If they have the permission,

the tool will state their identity, time of

entry, and activities conducted. If they

lack the authority to do so, the machine

will make a report of an attempted

breach of security (Isensee, Teuteberg,

Griese, & Topi, 2020, p. 2). Relevant

authorities can then investigate the issue

immediately. This technology is meant to

eliminate or significantly reduce cases of

theft of goods on transit or those that are

already in the warehouse.

Figure 3.2: ML techniques relevant in

logistics management

Artificial Intelligence Opportunities and

Benefits for MSMEs in the Logistics

Sector

The concepts of AI and ML are rapidly

gaining relevance and acceptance in the

field of business. According to Belhadi et

al. (2021, p. 10), it has become evident

that AI has a unique capacity that is

beyond that of humans when it comes to

making accurate data-based predictions.

The machine can process Big Data within

a relatively short period, and then use it

to address a specific problem that is

affecting a given business entity. Kersten

et al. (2019) state that AI has immense

opportunities and benefits for micro,

small, and medium business entities (p.

49). They only need to understand how to

correctly apply it in specific operations of

their businesses. It is necessary to discuss

the specific opportunities and benefits

that AI presents to these firms.

AI Opportunities for MSMEs in the

Logistics Sector in Logistics Sector

MSMEs face some unique challenges,

which make it necessary for them to find

ways of competing favorably against

some of the larger competitors. For a firm

to succeed in the integration of AI into its

operations, it should understand the

opportunities that the new technology

presents and how it can fully take

advantage of them. The following are

some of the specific opportunities that

using this technology presents to a firm.

New strategies of reaching out to new

suppliers. One of the unique

opportunities that AI presents to MSMEs

in the SCM is the ability to identify, select,

and reach out to new suppliers.

According to Singh et al. (2012), one of

the most important factors that a firm

has to consider is effective ways of

identifying suppliers and communicating

with them effectively (p. 171). Changes in

the market make it necessary for a firm to

regularly evaluate suppliers that can offer

the best deals in the market. Using AI, a

small or medium company can process

large volumes of data within a short

period and select suppliers than offering

the best deals in the market. It can easily

conduct a comparative analysis to

determine how it can get the best value

for every purchase that it makes.

Improved processing of clients’ orders.

According to Klumpp and Ruiner (2018),

large organizations handle numerous

transactions every one hour as they

struggles to serve its customers in the

most efficient way possible (p. 7). Such a

large volume of a transaction within that

short period cannot be possible if

technology is not applied. Small and

medium enterprises must understand

the significance of AI in making such large

volumes of transactions possible. Using

AI, the logistics department can easily

interact with the sales department,

process clients’ orders, and facilitate

their delivery within a short time.

The technology creates a unique

opportunity for expansion for these small

companies. Instead of hiring more

employees to process customers’ orders,

the task is delegated to the machines. It

means that the firm will have an

increased capacity in terms of processing

clients’ orders, without necessarily

increasing the size of its workforce.

Belhadi et al. (2021, p. 20) also reiterate

that customers can track their orders

once they make their payments. Instead

of constantly calling the firm to make

such inquiries, they can easily use their

mobile phone apps or personal

computers to know when their products

will arrive. Such services tend to increase

customer satisfaction, which creates

further opportunities for growth.

Better ways of exploring new markets.

One of the primary goals of every MSME

is to achieve consistent and sustainable

growth in a highly competitive business

environment. However, expanding to

new markets beyond the primary or

understanding future market trends one

often presents numerous challenges, one

of which is how to manage the supply

chain in the new market (Pournader et

al., 2021, p. 16). AI can help address such

a problem when correct data is collected

and processed. AI can provide various

simulations about possible logistics

strategies that can be used in the new

market. The machine will collect and

analyze data about the new market, and

then propose ways in which the firm can

address specific challenges in the market.

If a firm in the United Kingdom or the

United States seeks to start operations in

China or India, it has to understand socio-

cultural, economic, and political forces in

the new country, and how they influence

a firm’s supply chain. Using AI, a firm can

easily assess these forces and determine

how they can be managed to ensure that

operations of the firm are not negatively

affected.

Ability to conduct detailed market

research. At the initial stages of AI

development, there was a general belief

that one had to educate the machine by

continuously feeding it the right data.

However, that has changed as technology

keeps advancing. AI has gained the

capacity to self-educate (Toorajipour et

al., 2021, p. 511). It can easily gather data

from the public domain, learn from it,

and make critical decisions and

predictions without any significant

human input. This new characteristic of

AI makes it an effective tool for

conducting market research. In SCM, one

only needs to specify the nature of the

research that should be conducted. The

machine is capable of gathering

information available in online databases

and answering specific questions of

concern to the firm. It can go a step

further to inform the management of the

best path to take based on several

simulations conducted. In such an

environment, the company will always

base its decisions on actual data instead

of personal feelings and rumors.

Identifying and eliminating wastage and

theft. One of the major challenges that

small and medium enterprises face is

pilferage when goods are in transit. Some

unscrupulous employees believe that the

best time of stealing from their employer

is when delivering materials to the firm or

when delivering finished products to

customers. The firm losses a lot of

resources when such cases of theft are

rampant. Using AI, the firm will have clear

data of materials purchased. The

machine can autonomously check if what

was delivered to the firm is consistent

with what was purchased (Belhadi et al.,

2021, p. 2). It can also accurately

determine if products leaving the firm are

those fully paid for by the client. The

strategy significantly reduces and in some

cases completely seals loopholes that

employees may be using to steal from the

firm.

The technology can also help in

eliminating wastes within the firm.

According to Aarstad and Saidl (2019),

automates the filing system, which

improves efficacy and eliminates

wastage. One of the main causes of waste

generation within a firm is offering excess

raw materials that cannot be consumed

within a specified period. When supplies

are in excess, a significant amount will be

wasted within the firm. It also encourages

theft, and the management may not even

realize that workers are stealing from the

firm. There is also the additional cost of

warehousing. Using AI, the firm will be

ordering only what is needed and

consumable within a specific period. It

will not only reduce the cost of

warehousing but also the wastage and

possible theft among employees. The

management can use the savings made in

facilitating the expansion of the firm to

new markets.

AI Benefits for MSMEs in the Logistics

Sector of MSMEs

AI remains to be relatively new

technology, and many business entities

are yet to fully integrate it into their

operations. Toorajipour et al. (2021)

believe that most executives are not yet

fully aware of the benefits that they may

get from it (513). They still look at AI as a

technology that is limited to data

processing and making predictions when

directly assisted by humans. However,

recent developments in this field have

empowered this technology, and its

capacity to make accurate predictions

and effective data-based decisions are

unmatched. Small and medium

enterprises can benefit significantly from

this technology if they apply it correctly in

supply chain management. The following

are the direct benefits of AI to MSMEs’

SCM.

Eliminating the need for large

warehousing units. One of the obvious

benefits of AI to small and medium

enterprises is that it eliminates the need

for large warehousing units at a firm.

MSMEs cannot afford to have large

warehouses to store raw materials or

finished products. As such, most of them

have been relying on a just-in-time (JIT)

system of inventory management to

ensure that the needed materials are

made available only when they are

needed (Kersten et al., 2019, p. 65). AI

improves this system by providing

relevant data. It can make an accurate

prediction about the flow of materials on

the production floor, making it possible

to know the specific time that the

procurement department needs to

purchase the relevant materials.

AI has also become a powerful tool in

making sales forecast in the market. This

capability makes it possible for it to have

accurate data about the entire flow of

inventory, from purchasing raw materials

from suppliers, moving them to the plant,

processing them, and finally releasing

finished products to the market. Having

an accurate demand forecast means that

the firm will only produce what the

market can consume within a given

period. As such, there will be a reduced

need to store finished products in the

warehouses.

Optimization and automation enhance

productivity. Supply chain management

is one of the critical functions that define

the success of a firm. MSMEs must

ensure that they source the relevant

materials in time to avoid cases where

production is stopped. Similarly, finished

materials have to be made available in

the market at the right time and volume.

Traditionally, decisions and predictions

would be made by humans (Toorajipour

et al., 2021, p. 506). The problem that

firms often face is that human error is

common when there the machine does

not support the process. Human bias and

limitations mean that they cannot always

be accurate as would be desired.

AI optimizes SCM activities through

automation, which enhances

productivity. Although humans are still

expected to be actively involved in the

process, the automation process makes

work easy. The production department

can easily monitor the flow of products

based on sales made within a given

period. As such, they can determine the

volume of products they need to make

available within a certain period. The

same information will be immediately

available to the procurement

department, which will be able to predict

when to make raw materials available to

the production unit. At the same time,

the finance department will be informed

about the needed resources in each of

these departments to facilitate smooth

and effective operations.

Real-time data facilitates real-time

supply chain management. One of the

benefits of using AI is that it makes it

possible for a firm to have real-time data.

Through machine learning, it is possible

to train AI to collect real-time data about

market demands and the availability of

specific materials needed for the

production process. Pournader et al.

(2021) explain that one challenge that

MSMEs face is that sometimes they use

obsolete data to make decisions (p. 41).

As such, they make errors because the

information upon which they base their

decision is not accurate.

AI can be trained to constantly collect

data in the market. It will provide

updated information, enabling the

management to accurate predictions.

With real-time data, the management

can make real-time supply chain

decisions. Christiansen and Skrinjaric

(2021) warn that when using AI to collect

data from the public domain, care should

be taken to identify and avoid misleading

or biased data (p. 39). It is also essential

for the user to ensure that third parties

are not allowed control of the AI because

they can manipulate it in a way that may

provide misleading information.

High business continuity is enhanced by

advanced monitoring and maintenance

possibilities. A significant number of

firms in the United States fail to reach the

third year of operations. One of the

primary factors that lead to the failure of

these start-ups is the inability of the

manager to effectively monitor and

control various operational activities. At

this initial stage, the proprietor is often

forced to hold several managerial

positions because of the few numbers of

workers. It is easy for the manager to fail

to maintain effective monitoring and

evaluation of employees’ work. AI makes

it possible for the proprietor to monitor

the performance of employees. AI also

enhances the capacity to monitor

inventory and how it is used within the

firm (Belhadi et al., 2021, p. 9). The

enhanced monitoring and maintenance

that is facilitated by this technology

minimizes wastage, cuts costs of

operation, and enhances efficiency in

supply chain management, which in turn,

enhances business continuity.

Improved and sustainable working

conditions. Procurement and logistics

can be challenging, especially when for a

firm in the manufacturing sector. In small

and medium enterprises, these tasks are

often assigned to a few employees

because of limited resources within the

firm. Having to collect data, process it,

and use the information to make critical

decisions and actions can be demanding.

AI helps in creating sustainable working

conditions for these MSMEs by

simplifying most of the procurement

tasks. The machine can collect and

process data, then use the information to

make accurate procurement decisions.

Employees can then focus on other tasks

that cannot be conducted by the machine

(Toorajipour et al., 2021, p. 510). The

reduced workload makes it possible for

employees to focus on innovative ideas in

their workplace.

Employees’ safety is one of the main

challenges that MSMEs face in their

supply chain management. Some of the

activities that workers engage in expose

them to dangers. The occupational health

and safety of employees are some of the

major requirements that the government

has set for firms irrespective of their size.

When a worker sustains an injury while

they are in their official duty, it is the

employer will be held fully liable. The

introduction of intelligent robots reduces

cases of such accidents happening. They

are less likely to make mistakes that may

cause an accident in the warehouse

(Klumpp & Ruiner, 2018, p. 5). It means

that the firm is less likely to face

litigations related to accidents in the

workplace environment.

Customization and personalization for

clients. Amazon.com and other large

corporations have learned the

importance of collecting information

from clients and then customizing and

personalizing products to meet their

specific needs. Alexa has become a

powerful tool that Amazon.com uses to

collect information from its clients to

understand their preferences. Small and

medium enterprises can use similar

technologies to understand the specific

needs of their clients. Some individuals

prefer having their products delivered at

their offices while others want the

product delivered at home or in social

settings at specific times. Understanding

these specific customers’ needs and

finding ways of meeting them can help a

firm to create a pool of loyal customers

(Belhadi et al., 2021, p. 9). Having such a

database, the logistics unit will have clear

information about the best time and

place of delivering customers’ orders.

The ability to maintain constant

communication with these clients also

helps in ensuring that cases of lost

products are eliminated.

Enhanced creativity and innovation. In

the current competitive business

environment, firms are under immense

pressure to find ways of meeting the

needs of customers in ways superior to

those of competitors. They also need to

introduce better ways of operations that

can cut the cost of operations to help

improve profitability without necessarily

increasing the product price. It means

that they have to constantly find unique

ways of operation. Creativity and

innovativeness can only be enhanced

when workers have some free time at

work to test and actualize new ideas. It

also thrives in an environment where

there is a steady flow of relevant and

updated data (Kersten et al., 2019, p. 75).

AI can be trained to undertake most of

the tasks in the logistics department.

While large computers are used to collect

and process data and use the information

to make complex predictions,

simulations, and predictions, robots can

be used to handle physical tasks in

warehouses. It means that enough time

will be available for employees to think

creatively and test their ideas to enhance

their productivity (Klumpp & Ruiner,

2018, p. 6). The processed information

made available by the AI also makes them

understand emerging needs in the

market, making it possible for them to

develop relevant concepts that the firm

needs to prosper.

Reduced human error. One of the biggest

promises of AI and ML is the elimination

of human errors. Ceyhun (2020) notes

that even when people are presented

with accurate data, they are likely to be

influenced by personal biases and

stereotypes when making decisions (p.

76). However, machines tend to be highly

objectives when presented with accurate

data. When properly trained, AI makes

very accurate forecasts and decisions

that can help a small or medium

enterprise to achieve rapid growth. Cases

of human error are sometimes caused by

the inability to collect and process data

accurately. This challenge is effectively

addressed by AI technology (Christiansen

& Skrinjaric, 2021, p. 121). In supply chain

management, employees can use AI to

select the best suppliers in terms of the

quality they offer and pricing. Accurate

simulations can also be conducted to

determine how to move products from

the supplier, store them within the firm,

and then deliver them to the market.

Elimination or a significant reduction of

human error lowers the cost of

operations while at the same time

increasing productivity and efficiency.

Reduced number of employees needed.

Small and medium enterprises cannot

afford to have a high number of

employees. However, they still have to

compete against large multinational

corporations that employ thousands of

employees. There is a need for these

firms to find ways of working effectively

with their limited number of workers. AI

makes it possible for these small firms to

operate successfully with a handful of

employees. Robots can now undertake

most of the physical jobs that were

previously conducted by humans in the

procurement sector (Toorajipour et al.,

2021, p. 505). On the other hand,

complex computers can be used to

collect and process data. Doing so makes

it possible for a firm to function efficiently

without the need to have a large number

of workers. The fact that machines can

operate for more than 24 hours without

the need for a break, as is always the case

with humans, means that a great amount

of work can be completed within a

relatively short period.

Enhanced consistency and reliability.

MSMEs need to be consistent and

reliable when meeting the needs of their

customers. When clients approve of a

specific quality of products, it is essential

to ensure that it is maintained until such

a time when it can be improved based on

changing tastes and preferences. AI

makes it possible for a firm to maintain

consistency in many ways. First, it will

ensure that information about the best

suppliers in the market is maintained.

Products can only be of the right quality

if the standards are met when sourcing

for raw materials. Secondly, AI can be

used to ensure that delivery methods are

standardized (Hellingrath & Lechtenberg,

2019, 91). It can also facilitate close

coordination among all the departments

within the firm to ensure that clients’

expectations are always met.

Training and collaboration. AI and ML

create an environment where it is easy

for a small or medium enterprise to

conduct training and enhance

collaboration among employees. AI

creates a system where a firm can easily

assess the knowledge gap of its

employees and job market requirements.

It means that targeted training, which

focuses on addressing the identified

gaps, is possible (Dash et al., 2019, p. 49).

AI significantly enhances data sharing

within departments of a firm and in the

entire organization. Sharing of data also

encourages collaboration among

workers. They can identify a specific

problem and then work closely to find an

effective solution. Targeted training and

collaboration among workers enhance

the efficiency of a firm and its ability to

achieve sustainable growth.

Environmental protection. Global

warming and climate change have

become major concerns in the

21st century. Many governments and

scientists are struggling to find ways of

reducing the emission of greenhouse

gases and eliminating harmful industrial

effluents. In the United States and many

other countries around the world, strict

policies have been introduced to ensure

that companies do not pollute the

environment. Some of these countries

have introduced limits of carbon

emissions that every firm cannot go

beyond within a specific period. AI

promises to help firms to operate

optimally without going beyond these

limits (Pournader et al., 2021, p. 13). The

use of robots that can use green energy

instead of heavy machines that use diesel

is one of the best ways of protecting the

environment. AI can also be used to

develop simulations of how a firm can

operate optimally with the least possible

production of industrial effluent and

greenhouse gases. The same technology

can also help in developing effective ways

of managing wastes. Figure 4.1 below

summarizes the main benefits of Industry

4.0 to small and medium enterprises.

Figure 4.1: Benefits of AI

Artificial Intelligence Challenges for

SMEs in the Logistics Sector

When embracing AI, micro, small, and

medium enterprises must understand

that there are some challenges that they

have to manage. The previous chapter

has outlined benefits and opportunities,

which make this tool an effective one in

enhancing supply chain management

among these entities. Holistic analysis

also focuses on challenges that exist, and

how they can be managed to ensure that

these firms can use them effectively. The

following are the primary challenges that

these firms have to put into

consideration.

Cyber-Security Issues

One of the major issues when using AI to

facilitate operations in the logistics sector

is cyber-security threats. A major cyber-

security threat is system manipulation. It

is a situation where the system is

deliberately fed with misleading and

malicious inputs. The intention is to

expose the machine to a picture that

does not reflect real-world forces.

According to Toorajipour et al. (2021), AI

and ML rely on data to learn about what

it is expected to do (504). It takes time for

the machine to learn about what it is

expected to do. The process involves

feeding it the correct data so that it can

make correct decisions and predictions

based on the information gathered.

Cyber-criminals can manipulate the

system by exposing it to wrong

information.

The outcome is that the decisions and

predictions that the AI makes will be

based on wrong information. Instead of

helping the logistics department in

making the right decisions, it will

misguide the whole process based on the

data it was fed. Data corruption, also

known as data poisoning, has become a

major problem in AI, especially when an

attacker has a specific malicious goal to

achieve, and easy access to the AI data

system (Belhadi et al., 2021, p. 8). A

competitor can easily attack a firm will

poisoned data to mislead them into

taking actions that do not reflect realities

in the market. MSMEs are at a greater risk

of being attacked because of their limited

financial power to establish strong data

security measures.

AI machines are trained to sometimes

gather data independently from the

public domain. As Dash et al. (2019, p. 49)

observe, they do not necessarily need to

rely on their users to feed them with the

data they need to make decisions and

predictions. The problem is that it

exposes them to misleading information

on online platforms. A significant amount

of information available online is biased,

misleading, and sometimes meant to

achieve the selfish goals of some

individuals or entities. When AI has

access to such data and fails to realize

that it is untrue, it may base its decisions

on the information gathered. In such a

case, the machine will end up serving the

interest of external parties, which may be

the competitor, instead of the user.

Kersten et al. (2019) also warn against the

possibility of adversaries retraining the AI

to meet their selfish goals (p. 82). In such

a case, the AI is gradually fed inputs that

shifts their focus from their primary

purpose. Eventually, they become

irrelevant tools that cannot perform

specific tasks assigned to them.

Data privacy is emerging as a major

concern when using artificial intelligence

and machine learning technologies. AI

relies on large volumes of data to

function appropriately. The data has to

be stored in a database to ensure that the

machine can have access to it whenever

it is necessary. The problem that arises is

that external parties may target the

database and extract information. In such

a case, the cybercriminal will easily

predict all the decisions that the machine

is likely to make in specific scenarios. In

such a case, the AI will be a major

weakness to the firm instead of being a

strength because its predictions will be

obvious to the rivals. Some private data

about suppliers or transporters may be

released to the public domain in such

attacks (Klumpp & Ruiner, 2018, p. 12).

The outcome will be a lack of trust with

important partners in the market.

It is crucial to find an effective way of

addressing these challenges. Toorajipour

et al. (2021) advise that when a small or

medium-sized enterprise decides to use

AI in its logistics department, one of the

factors that have to be considered is the

protection of data from cybercriminals

(p. 515). AI and ML highly rely on data

that they have access to when making

crucial decisions and predictions. A firm

must understand that rivals can easily

target the machine and manipulate it to

their benefit. As such, it is essential to

protect the system from any form of

attack. If it is necessary to allow the AI

system to have access to information in

the public domain, some measures may

need to be taken to protect it from

malicious data. The system can be

protected by directing it to specific

websites while limiting its access to

others that are not credible.

A firm must ensure that it is always in

control of the AI system. Measures

should be taken, using the latest data

protection software, to ensure that any

breach is detected as soon as it occurs,

and appropriate measures taken to

address the identified problem.

Christiansen and Skrinjaric (2021)

recommend the use of resilient cyber

security solutions that can provide end-

to-end data encryption and protection,

such as Hardware Security Module (p.

144). They significantly reduce the threat

of external parties having access to the AI

system, including its database.

Computing Power

Artificial intelligence and machine

learning are rapidly gaining popularity in

many parts of the world. However,

Kersten et al. (2019) admit that AI and ML

are relatively new concepts (p. 39).

Although many people in the business

sector have heard about them, their

applications remain a mystery to many

people. The main problem is that

implementing AI in a firm requires

intensive training and a significant

investment in Big Data and related

technologies. A firm must be willing to

spend time and financial resources to

take their employees through rigorous

training on how to feed data to the

system and train it to make critical

decisions and predictions. They should

also learn how to work alongside the

machines without having any conflicts

and making important decisions.

Large multinational corporations have

the financial capacity to embrace AI with

ease. They can spend money on research

to understand how to use the new

technologies in its logistics sector.

However, the same cannot be said about

micro, small, and medium enterprises.

These small firms are struggling to meet

various financial needs in their

operations (Hellingrath & Lechtenberg,

2019, p. 49). They may not afford the high

initial cost of installing the system. They

may also lack the capacity to equip their

employees with the computing skills

needed to operate the system. The

option that they may have is to hire

individuals who already have the needed

skills. The challenge that they are likely to

face is that some employees may

demand higher salaries because of their

unique skills.

The MSMEs must find ways of

overcoming this challenge because

ignoring AI and ML is not an option. One

of the best ways of addressing this

challenge is to have a gradual process of

embracing the new technology. It can

select a few loyal and techno-savvy

employees to learn more about AI and

ML, and how they can be applied in a

normal operation of a business (Dash et

al., 2019, p. 48). Instead of making a

heavy initial investment into the system

at a time when the firm is financially

strained, it can make small steps towards

understanding and embracing the

system. One of the best ways of achieving

the intended goal is to embrace best

practices in the industry. Instead of

spending a lot of time and resources

trying to define how to use this new

technology, these small firms can use

models that have already been

developed by large corporations.

Poaching talented employees with skills

in AI and ML may be another way through

which they can enhance their computing

power.

Trust Deficit

Trust is a major concern among business

executives who are interested in

embracing this new technology. AI and its

capabilities are relatively new, and some

still believe that they cannot fully trust it

to make decisions (Toorajipour et al.,

2021, p. 511). A section of business

executives considers AI to be too complex

to be applied successfully in the business

sector. Its success in the security and

health sectors has not convinced these

stakeholders. Some of them argue that

they do not understand how employees

can easily and successfully interact with

AI and its levels of automation without

causing conflicts. As such, those who are

expected to lead and facilitate

automation have expressed their lack of

trust towards the new system. The

resistance has had a major challenge in

the implementation of AI in the supply

chain sector, especially among MSMEs.

When the management and employees

feel that they cannot trust the system to

perform its functions effectively, they are

less likely to support its implementation.

The current problem of lack of trust

among some stakeholders should be

addressed to ensure that AI and ML are

embraced as tools that can enhance

operations in the logistics sector. Kurt

Lewin’s change model can be used to

help address this problem. This model

identifies three stages of introducing

change within an organization. The first

stage, known as unfreezing, involves

preparing the entire team for the change.

They need to understand the weaknesses

of the current system, and how the

proposed change will help solve the

problems. It also involves undertaking

the necessary training so that the team

can understand what is expected of

them. The next step is to undertake the

actual change. The team is expected to

use the already established plan to

introduce the relevant change. The last

stage involves refreezing. The team will

focus on assessing the new system,

identifying new challenges, and then

finding solutions to them. Using this

model will help eliminate possible

resistance to change by ensuring that

everyone is adequately prepared for it.

Figure 6.1. Kurt Lewin’s change model

Limited Knowledge

AI and ML have a huge potential that

MSMEs should fully take advantage of to

manage stiff competition in the market.

However, a major challenge arises from

the fact that there is limited knowledge

about this emerging technology (Aarstad

& Saidl, 2019, p. 134). The first challenge

that most small businesses face is that

their executives have limited knowledge

about it. They might have heard about AI

and ML, but they lack detailed knowledge

about their relevance in business

settings. The controversies about this

technology witnessed in the recent past,

especially on its capacity to replace

humans in decision-making, have also

caused confusion among these

executives, making it difficult for them to

embrace it.

Most of the employees who are already

in the job market have limited knowledge

about artificial intelligence. They may be

having the interested to know more

about it and how it can be implemented.

However, they lack the academic

background to fully understand the

system. As such, some may develop the

fear that instead of AI making their work

easy, it can potentially eliminate the need

for their services in their current firm

(Hellingrath & Lechtenberg, 2019, p. 53).

Another concern is that employing a large

number of fresh college graduates with

some background training in AI would

mean trading experience and loyalty for

one aspect of technology.

Micro, small, and medium enterprises

must find ways of addressing such

challenges. According to Toorajipour et

al. (2021), the number of institutions of

higher learning that offer training in AI is

increasing (p. 510). This technology is

gaining relevance in various fields and

many colleges currently offer some form

of training related to it. Firms should

consider enrolling their current

employees in these colleges so that they

can sharpen their skills in this field.

Executives can also enroll themselves in

online classes to enable them to enhance

their understanding of this new

technology. When hiring, a firm can

emphasize the need for the recruits to

have some basic skills in AI and ML.

The Problem of Bias

AI machines learn from the data to which

they have access. One of the main

challenges that an organization may face

when using AI in the supply chain is bias

(Kersten et al., 2019, p. 57). Sometimes

the system is exposed to bias against or

towards some people or some things. It is

common to find cases where many

organizations prefer hiring male

employees for engineering jobs while

females are given priority for secretarial

or marketing jobs. There is also a

common bias in the United States and

other parts of the world in favor of whites

at the expense of other races, especially

blacks. When AI is exposed to such data

during the learning process, it will base its

decisions based on such bias.

In logistics, a firm may prefer to source

materials from specific individuals or

regions for particular reasons. Dynamics

in the industry creates a situation where

some of the best suppliers or

transporters today may not be the best in

the future (Dash et al., 2019, p. 50).

However, when the machine is exposed

to such bias, they may continue making

choices, which may not be the best for

the firm. In such a case, instead of

facilitating optimum decision-making, it

will become a misleading tool that causes

loss to the firm.

It is essential to find a way of addressing

the problem of biases of AI to ensure that

it remains as effective as possible. One of

the best ways of doing so, as Klumpp and

Ruiner (2018) note, is to ensure that the

AI is exposed to unbiased data during the

training process (p. 2). The user must

ensure that the machine learning process

is free from any racial, religious, or

gender biases. Continuous learning is

another essential tool that can be used to

help the machine make optimal decisions

and accurate decisions. As the forces in

the market continue to change, the data

that the machine uses should also be

updated accordingly. Data on factors

such as the best suppliers, reliable

transporters, and security issues should

be regularly updated to enhance the

efficiency of these machines.

Infrastructure

The infrastructure needed to have a fully

functional AI system in the logistics

department may be a challenge to micro,

small, and medium enterprises. A firm

needs to have large and secure databases

to facilitate the creation of Big Data. It will

then require an AI system that is

specifically designed to facilitate

operations in the logistics department

(Ceyhun, 2020, p. 119). For large

corporations, developing such an

infrastructure may not be a major

challenge because of their financial

capacity. However, the same cannot be

said about smaller firms with a small

budget. Sometimes they may be forced

to choose between installing the

infrastructure or facilitating expansion or

hiring more employees. The decision will

always be based on what the

management considers more urgent and

important to the firm. Sometimes having

the AI infrastructure may not be

considered a priority when the firm is

faced with a financial constraint. Singh et

al. (2012) also warn against poor AI

systems that cannot function efficiently

as expected (p. 172). Some of these AI

systems are developed by individuals

who are still under training, which means

that their creation may have several

weaknesses.

The cost of installing an AI infrastructure

for the logistics department may be

significantly high. However, the

technological trends witnessed in the

recent past show that firms, irrespective

of their size, cannot avoid embracing

emerging technologies (Toorajipour et

al., 2021, p. 504). It means that these

MSMEs must find ways of overcoming

the challenge of having the right

infrastructure for their logistics

department. Gradual planning for AI

infrastructure development may be one

of the ways of overcoming the challenge.

The company can set aside a specific

amount of its monthly income towards

purchasing the infrastructure. It means

that its installation can be categorized as

a mid-term plan that should be realized

within a year. Care should be taken when

selecting the vendors who are selling the

relevant system. It is advisable to select

established vendors with a good

reputation to avoid the problem of

having an ineffective AI system.

Integration of AI into Existing Systems

When the management of a small or

medium enterprise decides to embrace

AI technologies, another challenge that it

has to deal with is integrated into an

existing system. The firm will have to find

a way of introducing the AI system into an

already existing system within the

logistics department. Klumpp and Ruiner

(2018) explain that the integration

process may face various challenges (p.

21). First, the team of employees may

lack the requisite skills needed in the new

platform. Secondly, some employees’

jobs may have to be redefined, especially

those whose roles will be taken by the

machine. There is a possible threat that

some of the employees may reject and

deliberately frustrate the introduction of

the new system if they feel that their

position within the firm will be

threatened. It is also possible that a firm

may have to make a radical shift from

traditional methods of operations to a

new one that is based on the new

technology.

The possible integration challenges will

need to be addressed to ensure that

these small firms can effectively embrace

AI in their supply chain management. To

address the challenge of inadequate skills

among employees, the firm will need to

train a section of the employees who will

be operating the AI in the logistics

department. They should be equipped

with relevant skills necessary for them to

feed and allow the AI systems to make

decisions and predictions (Dash et al.,

2019, p. 48). The firm may also find it

necessary to recruit new employees with

the right skills. When redefining the roles

and responsibilities of employees in line

with the changes introduced, the

management should ensure that it is

done in a way that does not cause fear

among workers. They should be informed

that the new system is meant to

empower them instead of eliminating

them from the firm. The goal is to ensure

that all the relevant stakeholders support

the new system. When it is necessary to

make a radical shift from traditional

methods to new ones, employees should

be adequately trained and prepared for

the change.

Inability to Reach Human-Level in

Decision-Making

One last challenge of AI in the logistics

sector for MSMEs worth discussing at this

stage of the paper is its inability to reach

human-level in decision-making.

According to Klumpp and Ruiner (2018),

there has always been a debate about

whether AI has the capacity of a human

when it comes to making rational

decisions (p. 26). Some AI enthusiasts

have even argued that these machines

are superior to humans when it comes to

making accurate predictions and data-

based decisions. It may be true that AI,

when fed with the right data, can make

accurate predictions and effective

decisions.

Sometimes their predictions may be

more accurate than that of humans

because of their capacity to process large

amounts of data within the shortest

period possible. However, they lack the

compassion that humans have, making

them relatively inferior (Singh et al.,

2012, p. 178). Sometimes it may be

necessary for a firm to make a decision

that is seen to protect humanity and the

environment as opposed to making

profits. However, these machines are

often trained to focus on maximizing

profits while at the same time minimizing

expenses and losses. Some of their

decisions may have serious negative

implications on a firm’s reputation

because of the lack of compassion.

Small and medium enterprises should

find a way of addressing this challenge. AI

has proven to be a highly sophisticated

system capable of doing what only

humans can do. However, they need to

be trained to do that which they are

expected to do. Toorajipour et al. (2021)

argue that it is possible to train the AI to

be compassionate when making

decisions (p. 514). Although these

machines lack human feelings and some

of their decisions may be considered self-

centered and profit-oriented, they are

still essential. To address this challenge,

the management can ensure that the

final decisions are made by humans. The

machine can be allowed to develop a

pathway to different possible actions.

The management can then understand

the possible outcome if a certain decision

is made. Using the information that the AI

has processed, the management can then

make decisions that reflect other human

needs the machine could have ignored.

Figure 6.2 below identifies some of the

major challenges that AI presents to

MSMEs when used in supply chain

management.

Figure

6.2. Challenges of AI in SCM

The Future of MSMEs’ Logistics in the

World of Artificial Intelligence

The last century has seen rapid growth in

popularity and application of AI in the

corporate world. According to

Christiansen and Skrinjaric (2021), the

gaming industry helped in testing the

efficiency of AI and ML. In the current

stage, technology has advanced into the

fields of healthcare, security, transport,

manufacturing, and communication

among others (p. 67). The need to use

real-time data has made it possible for

companies to use this technology in

making important decisions that define

their operations. As Ceyhun (2020) puts

it, the world is moving towards a

corporate world that is fully dependent

on artificial intelligence (p. 183). It is

becoming critical for companies to make

decisions that are based on data to avoid

mistakes and misinformation.

Micro, small, and medium enterprises

have to embrace AI technology in their

operations. As the competition stiffens,

these firms find themselves in positions

where they have to be accurate in making

their predictions. They are moving into a

world where every decision that they

make has to be based on accurate data. It

starts by investigating and determining

specific customers’ needs in the market

(Hanne & Dornberger, 2017, p. 85). The

challenge that exists is that these needs

may change depending on various forces

in the market. The fact that these MSMEs

are not industry leaders means that they

may not dictate most of these changes.

Once customers’ needs are understood,

the firm must determine the demand

size. The information will then be used to

inform the production process. To help

lower the cost, improve quality, and

enhance productivity, a firm must find

ways of embracing smart production. AI

makes it possible for these small firms to

embrace these new methods of

manufacturing.

In the logistics sector, it is becoming

impossible for small and medium

enterprises to ignore AI and ML. The

numerous benefits that the new

technology offers have been discussed in

the previous chapters of the paper. The

current trend shows that firms can no

longer avoid this technology. One of the

reasons why it is currently unavoidable is

the need to deliver products in the

market or to the manufacturing plants in

time (Klumpp & Ruiner, 2018, p. 3). Using

AI, drivers can know the best routes that

they can take to reach their destination

within the shortest time possible. Such

high levels of efficiency are only made

possible through the use of AI. There is

also the need to cut the logistics cost of

operations. Besides identifying the best

routes that are cost-effective for a

company’s trucks, AI also helps to fight

theft and reduce the perishability of

products. All these factors help in

lowering the overall cost of operations.

As such, any MSME that is keen on

achieving sustainable competitiveness in

the market needs to embrace the

technology.

It is common for employees and other

stakeholders to resist change, especially

when they feel that their current position

will be threatened. The management of

these small and medium enterprises

must realize that their employees may

not support the introduction of AI in the

logistics sector because it may not only

replace junior officers but also mid to top

management officers because of its

capacity to make accurate predictions

and data-based decisions (Singh et al.,

2012, p. 175). Having en effective

integration of AI is essential in eliminating

such resistance. Lewin’s change model,

discussed above, may help in introducing

change in these organizations.

The Relationship between ML Adoption

in Logistics and MSMEs Performance

Machine learning is gaining popularity

among small and medium enterprises

keen on redefining their logistics

operations. Many firms have realized that

they have to digitize their logistics as a

way of remaining competitive. The new

technology enables these firms to make

critical decisions based on accurate

predictions and scientific simulations

(Isensee et al., 2020, p. 6). It helps in

eliminating human errors, which may

significantly affect the ability of a

company to meet customers’ needs in

the best way possible. In this section of

the report, the focus is to discuss the

relationship that exists between ML

adoption in logistics and MSMEs’

performance. Figure 3.3 below identifies

specific performance-related factors in

logistics.

The performance of individual drivers will

be significantly enhanced when ML is

introduced in logistics management. The

performance of a firm directly depends

on the performance of an individual

employee. In logistics, drivers play a

critical role in ensuring that goods are

delivered at the right time and without

damage or theft. ML technology makes it

possible for drivers to be efficient in their

operations because of the assistance

from deep learning and text/image

processing technologies (Pournader et

al., 2021, p. 11). Driver-assist technology

helps drivers to make critical decisions

while they are on the road. They can

identify routes that may have traffic

congestion and the technology can help

suggest alternative routes. As such,

delays that may disrupt normal

operations within the firm or customer

dissatisfaction are eliminated. Such

delays may also affect the quality of

goods on transit when handling

perishable products (Belhadi et al., 2021,

p. 7). The driver-assist technology makes

it possible for the pilot, captain, or driver

to monitor the conditions of products on

transit and to set optimal temperature

and moisture that can help preserve

them in the best way possible while on

transit.

Autonomous driving remains one of the

most promising opportunities that AI and

ML will offer to the logistics sector. For

small and medium enterprises, this

technology will offer a perfect

opportunity to scale up operations and

enhance productivity. Current

development in autonomous vehicles,

especially the one which is headed by

Tesla, strongly indicates that driverless

trucks will soon be a reality (Min, 2010, p.

16). If this is achieved, MSMEs will be able

to facilitate the transportation of large

volumes of products within a short

period. Machines do not need to have a

break in ways that humans do. They

cannot complain or demand additional

pay when they have to work for 24 hours

a day. It means that a firm can easily

enhance its productivity in the logistics

department when this technology is

applied.

Cloud computing is another major

opportunity for growth that AI and ML

offer to small and medium corporations.

In logistics management, data has

become a critical factor that defines a

firm’s success. Making decisions based on

real-time data is essential, especially

when handling perishable products. A

firm needs to have accurate information

of demand in the market and other

factors to know the number of products

that should be made available (Kersten et

al., 2019, p. 32). Cloud computing makes

it possible for these firms to access,

process, and utilize large amounts of

data. Instead of relying on guesses and

unfounded speculations, these entities

can streamline their logistics operations

based on real-time data.

Cloud computing also facilitates effective

communication among stakeholders in

value chain management. Once the

demand is determined by the marketing

department, the manufacturing unit will

plan its activities properly and accurately

make an order of materials needed. The

information will be relayed to the

suppliers through the procurement

department. The logistics unit will then

have clear information on the amount

and nature of products that need to be

transported from suppliers to the firm

and from the firm to the market within a

specific period (Pournader et al., 2021, p.

17). The increased coordination and

cooperation of these departments and all

the relevant stakeholders improves the

performance of the logistics department

and the entire firm.

Driver risk assessment is another way in

which a small or medium enterprise can

enhance its performance in the logistics

sector. Aarstad and Saidl (2019) explain

that driver risk assessment involves

determining the likelihood of a driver

being in an accident while on official duty

(p. 43). Various factors may cause an

accident while a driver is on official duty.

They include fatigue that impairs the

judgment of a driver when they are

overworked, vehicles being in a state of

disrepair, poor weather conditions, and

poor conditions of roads, limited training,

drunk driving, and carelessness or

recklessness of a driver. Some of these

factors are direct human errors while

others are forces beyond the control of

the driver.

An MSME can use ML to identify all the

possible causes of accidents by using

driver risk assessment. Factors such as

the poor state of the cars and limited

training can be directly addressed by the

management of the logistics department.

Poor weather and dilapidated road

networks are beyond the control of a firm

or its employees. However, drivers can be

trained on how to maneuver in those

harsh conditions. Problems that are

directly related to irresponsive driving

can be addressed by the human resource

department through strict policies that

guide the actions and decisions of drivers

while they are on official duty. It

significantly enhances supply chain

resilience (Belhadi et al., 2021, p. 7).

These measures help in enhancing the

overall performance of the logistics

department.

AI and ML technologies make it possible

for the procurement and logistics

manager to conduct real-time monitoring

of drivers. Small and medium-sized

enterprises face numerous challenges in

supply chain management (Klumpp &

Ruiner, 2018, p. 4). One of the challenges

that small and medium enterprises

sometimes face is having rogue and

irresponsible drivers. Some of them

engage in personal business using

company trucks during their official

working hours. Others use different

excuses to justify delays caused by their

laziness and irresponsible behavior. AI

and ML offer a perfect solution to the

challenge. It makes it possible to know, in

real-time, the physical location of a

driver. They can easily determine if it is

true that there is a traffic jam or any

other challenge that may affect the

smooth flow of the trucks (Kersten et al.,

2019, p. 66). The technology also makes

it impossible for a driver to use a

company truck to run personal errands

because the information will be relayed

to the management immediately. As

such, ML makes drivers more

responsible, which in turn, enhances the

overall performance of the logistics

department.

Figure 3.3: AI in smart manufacturing

Conclusion and Recommendations

The digital transformation that has been

facilitated by Industry 4.0 has had a major

impact on how firms conduct their

operations. As shown in the discussion

above, data has become a critical factor

that defines the ability of a firm to

achieve success in the market.

Traditionally, managers would make

predictions and decisions based on data

currently available, which in some cases

would be outdated. However, AI and ML

have created a system that makes it

possible for companies to collect real-

time data and then use it to make

decisions and predictions. The

technology not only facilitates the

collection of data but also its speedy

analysis and usage in making critical

decisions within a firm. Unlike humans

whose judgments may sometimes be

clouded by misconceptions and personal

biases, AI tends to be highly objective,

basing its decisions and predictions on

data. As such, the technology has

surpassed human potentials in making

accurate predictions and developing

simulations that can yield the best

results.

The superiority of AI in managing data

and using it in making accurate

predictions has made it an efficient tool

in logistics management. The paper

shows that micro, small, and medium

enterprises can use this technology to

improve their supply chain management.

The technology enhances

communication among all stakeholders

in the procurement and logistics sector.

The management can easily track the

movement of goods on transit and

respond to emergencies that may arise

more effectively. Real-time data-based

decision-making and predictions help to

eliminate the challenge of purchasing

excess or limited raw materials that

inflates the cost of warehousing or

disrupts operations. It also ensures that

the firm makes available to the market,

the exact amount of products that are

needed. Industry 4.0 offers these firms a

unique growth opportunity, especially

when they are operating in highly

competitive markets dominated by large

multinational corporations.

Despite the obvious benefits of AI to

MSMEs’ SCM, the paper identified some

challenges that may have a significant

impact on these firms if they are not

effectively managed. Cyber-attack

emerged as one of the possible threats

that these firms may face when they fully

embrace AI. When external stakeholders

with malicious intentions take control of

these machines, they can manipulate

them. As such, they may become a

liability to a firm because of their

misleading predictions and decisions.

Computing power, trust deficit, limited

knowledge of the technology, the

problem of bias when training the

machine, infrastructure cost, integration

of AI into existing systems, and inability to

reach human-level in decision-making

were the other major concerns identified.

MSMEs must find ways of managing

these challenges to enable them to

embrace AI and ML in their SCM to

improve their performance and

productivity. The following

recommendations should be considered:

• A firm should conduct thorough

training for its employees before

introducing AI and ML technologies.

The goal is to ensure that they are

fully prepared once the technology is

introduced.

• MSMEs should invest adequately in AI

infrastructure to ensure that data can

be processed effectively and

disseminated to all the relevant

stakeholders within the firm.

• The management should put in place

effective cyber security measures that

will ensure that there is no breach.

Data that the AI has access to should

also be assessed to ensure that it is

accurate.

• While the machine can make

predictions and simulations, humans

should have the final decision on the

path that should be taken. As such, AI

will help them make information-

based decisions.

• Future research should investigate

dangers that MSMEs may face when

they fully rely on AI in their supply

chain management.

• It is also necessary for future scholars

to determine reasons why some of

the companies, irrespective of their

size, have deliberately avoided

embracing AI in their SCM.

References

Aarstad, A., & Saidl, M. (2019). Barriers to

adopting AI technology in SMEs.

Copenhagen, Denmark: Copenhagen

Business School.

Adixon, R. (2019). Artificial intelligence

opportunities & challenges in

businesses. Web.

Alfoqahaa, S. (2018). Critical success

factors of small and medium-sized

enterprises in Palestine. Journal of

Research in Marketing and

Entrepreneurship, 20(1), 4-16.

Amblee, R. (2018). The ugly fight:

Unleashing artificial intelligence against

global warming. Dubai: Gloture Books.

Anyoha, R. (2017). The history of artificial

intelligence. Web.

Belhadi, A., Mani, V., Kamble, S., Khan, S.,

& Verma, S. (2021). Artificial intelligence-

driven innovation for enhancing supply

chain resilience and performance under

the effect of supply chain dynamism: An

empirical investigation. Annals of

Operations Research, 3(1), 4-11.

Ceyhun, G. C. (2020). Handbook of

research on the applications of

international transportation and logistics

for world trade. Hershey, PA: IGI Global.

Chin, T., Hamid, A., Rasli, A., & Baharun,

R. (2012). Adoption of supply chain

management in SMEs. International

Congress on Interdisciplinary Business

and Social Science, 65(2), 614-619.

Christiansen, B., & Skrinjaric, T.

(2021). Handbook of research on applied

AI for international business and

marketing applications. Hershey, PA:

Business Science Reference.

Chung, S. H. (2021). Applications of smart

technologies in logistics and transport: A

review. Logistics and Transportation

Review, 153(1), 55.

Dash, R., McMurtrey, M., Rebman, C., &

Kar, U. (2019). Application of artificial

intelligence in automation of supply

chain management. Journal of Strategic

Innovation and Sustainability, 14(3), 43-

50.

Devang, V., Chintan, S., Gunjan, T., &

Krupa, R. (2019). Applications of artificial

intelligence in marketing. Economics and

Applied Informatics, 25(1), 28-36.

Didonet, S., & Díaz, G. (2012). Supply

chain management practices as a support

to innovation in SMEs. Journal of

Technology Management & Innovation,

7(3), 91-106.

Earley, S. (2014). The digital

transformation: Staying Competitive. IT

Professional, 16(2), 58-60.

Giudice, M., Scuotto, V., Garcia-Perez, A.,

& Petruzzelli, M. (2019). Shifting wealth II

in Chinese economy. The effect of the

horizontal technology spillover for SMEs

for international growth. Technological

Forecasting and Social Change, 145(1),

307-316.

Hanne, T., & Dornberger, R.

(2017). Computational intelligence in

logistics and supply chain management.

New York, NY: Springer Verlag.

Hellingrath, B., & Lechtenberg, S.

(2019). Applications of artificial

intelligence in supply chain management

and logistics: Focusing onto recognition

for supply chain execution. Cham,

Switzerland: Springer.

I-SCOOP. (2021). Industry 4.0 and the

fourth industrial revolution

explained. Web.

Isensee, C., Teuteberg, F., Griese, K., &

Topi, C. (2020). The relationship between

organizational culture, sustainability, and

digitalization in SMEs: A systematic

review. Journal of Cleaner Production,

275(1), 1-11.

Jacobs, T. (2020). Artificial intelligence

(AI) in supply chain & logistics supply.

Web.

Jovanovski, B., Seykova, D., Boshnyaku,

A., & Fischer, C. (2019). The impact of

Industry 4.0 on the competitiveness of

SMEs. Industry 4.0, 4(5), 250-255.

Kersten, W., Blecker, T., & Ringle, C. (Ed.).

(2019). Artificial intelligence and digital

transformation in supply chain

management: Innovative approaches for

supply chains. Proceedings of the

Hamburg International Conference of

Logistics (HICL), No. 27. Berlin, Germany:

GmbH.

Klumpp, M., & Ruiner, C. (2018).

Regulation for artificial intelligence and

robotics in transportation, logistics, and

supply chain management: Background

and developments. Network Industries

Quarterly, 20(2), 3-7.

Kumar, S. (2019). Advantages and

disadvantages of artificial

intelligence. Web.

Lawson, D. (2021). How AI is improving

supply chain management

processes. Web.

Li, B., Hou, B., Yu, W., Lu, X., & Yang, C.

(2017). Applications of artificial

intelligence in intelligent manufacturing:

A review. Frontiers of Information

Technology & Electronic Engineering,

18(1), 86-96.

Linh, N. K., Kumar, V., & Ruan, X. (2019).

Exploring enablers, barriers and

opportunities to digital supply chain

management in Vietnamese

manufacturing SMEs. International

Journal of Organizational Business

Excellence, 2(2), 101-120.

Madanchian, M., Hussein, N., Noordin, F.,

& Taherdoost, H. (2015). The role of SMEs

in economic development: Case study of

Malaysia. International Journal of

Academic Research in Management, 4(3),

77-84.

Matt, D. T., Modrák, V., & Zsifkovits, H.

(2020). Industry 4.0 for SMEs: Challenges,

opportunities and requirements. New

York, NY: Springer International

Publishing.

Min, H. (2010). Artificial intelligence in

supply chain management: Theory and

applications. International Journal of

Logistics Research and Applications 13(1),

13-39.

Pervaiz, S. (2020). The role of artificial

intelligence in supply chain management.

(Master’s Thesis, University of Applied

Sciences, Berlin, Germany). Web.

Pournader, M., Ghaderi, H.,

Hassanzadegan, A., & Fahimnia, B.

(2021). Artificial intelligence applications

in supply chain

management. International Journal of

Production Economics, 241(11), 7-16.

Rabayah, K. (2013). Adoption patterns of

information and communication

technologies by enterprises in developing

countries: Case study Palestine. Journal

of Information Technology in Developing

Countries, 23(3), 7-11.

Sendler, U. (2017). The internet of things:

Industrie 4.0 unleashed. New York, NY:

Springer International Publishing.

Singh, A., Wiktorsson, M., & Hauge, J. B.

(2021). Trends in machine learning to

solve problems in logistics. Procedia CIRP,

103(1), 67-72.

Singh, R., Kumar, R, & Shankar, R. (2012).

Supply chain management in SMEs: A

case study. International Journal of

Manufacturing Research, 7(2), 165-178.

Stinson, E. (2021). How artificial

intelligence (AI) can be used in supply

chain management. Web.

Tirkolaee, E. B., Sadeghi, S., Mooseloo, F.

M., Vandchali, H. R., & Aeini, S. (2021).

Application of machine learning in supply

chain management: A comprehensive

overview of the main

areas. Mathematical Problems in

Engineering, 20(1), 1-14.

Toorajipour, R., Sohrabpour, V.,

Nazarpour, A., Oghazi, P., & Fisch, M.

(2021). Artificial intelligence in supply

chain management: A systematic

literature review. Journal of Business

Research, 122(1), 502-517.

Wright, J. & Recht, B.

(2022). Optimization for data analysis.

Cambridge, UK: Cambridge University

Press.

Yuanita, N. (2019). Competition and bank

profitability.

Students also viewed