Understanding Artificial Intelligence Adoption in Operations
Management – Insights from the review of academic literature and
social media discussions
Purva Grover
Information Systems/IT area, Indian Institute of Management Amritsar, India
Email: groverdpurva@gmail.com
Arpan Kumar Kar
Information Systems area, DMS, Indian Institute of Technology Delhi, India
Email: arpan_kar@yahoo.co.in
Yogesh K. Dwivedi
Emerging Markets Research Centre, School of Management, Swansea University, Wales, UK
Email: y.k.dwivedi@swansea.ac.uk
Abstract
In this digital era, data is new oil and artificial intelligence (AI) is new electricity, which is needed in
different elements of operations management (OM) such as manufacturing, product development,
services and supply chain. This study explores the feasibility of AI utilization within an organization
on six factors such as job-fit, complexity, long-term consequences, affect towards use, social factors
and facilitating conditions for different elements of OM by mining the collective intelligence of
experts on Twitter and through academic literature. The study provides guidelines for managers
for AI applications in different components of OM and concludes by presenting the limitations of
the study along with future research directions.
Keywords
Artificial Intelligence; Technology Adoption; Model of AI utilization; Operation Management
1. Introduction
In the current era of digital transformation, data is considered as the new oil and Artificial
Intelligence (AI) is often perceived to be the new electricity which can create value out of this oil.
Warner and Wäger (2019) had defined digital transformation as the use of new digital technologies
for improving businesses by enhancing customer experience, optimizing operations, creating new
business models and many more. These new technologies used in digital transformation can be
cloud, blockchain, mobile, AI, Internet of Things and other smart technologies (Duan et al., 2019;
Dubey et al., 2020; Hughes et al., 2019; Ismagilova et al., 2019; Wamba and Queiroz, 2020). The
usage of technology for digital transformation is determined by organization’s attitude towards the
technology, perceived usefulness and perceived ease of use (Berlak et al., 2020; Grover et al., 2019).
Organizations has been significantly engaged in digital transformation (Burton-Jones et al., 2020).
AI and big data together are shaping economic, social and political spheres (Duan et al., 2019;
Dwivedi et al., 2019; Elish and Boyd, 2018; Wamba et al., 2015; Wamba et al., 2017). AI had been
defined as the system’s ability to interpret and learn from the digital traces (Haenlein and Kaplan,
2019). Metcalf, Askay and Rosenberg (2019) believes AI can amplify employee’s intelligence. AI
help employees in overcoming complex situation by presenting diverse and different solutions
(Jarrahi, 2018), and subsequently can provide prescriptive inputs in decision making process (Bader
and Kaiser, 2019). Employees should focus more on creative work and should learn how to
effectively use machines for mundane tasks (Jarrahi, 2018). Morikawa (2017) had pointed firms
having highly educated employees and having businesses worldwide expect that AI technologies
will have positive impact on businesses.
OM has been defined in literature as end-to-end organisational management activities and service
chains (Karmarkar and Apte, 2007; Subramanian and Ramanathan, 2012) which comprises of
several activities such as product designing, process designing, production of goods, planning,
scheduling (Zhao et al., 2020), personalized targeting, delivers, customizations, logistics,
outsourcing, and many more. The first research gap for the study is the gap pointed by Brock and
Wangenheim (2019) that managers have very less knowledge on how to use AI in their
organization’s operations. Therefore, this article presents usage of AI in different elements of OM
such as manufacturing, product development, services and supply chain.
The second research gap identified for the study are based on the gaps highlighted by Gunasekaran
and Ngai (2012), that there is a need to develop OM models for synthesising and converting
information into knowledge. Therefore, this study tries to explore the prospect of converting
information into knowledge by using AI on the data and information assets stored within the
organization, obtained from digital transformation initiatives. The third gap identified for the study
is the open question highlighted by Haenlein and Kaplan (2019), how people and AI supported
systems can peacefully coexist with each other. Therefore in this study, eight scenarios in the form
of propositions had been explored where authors feel employees and AI powered systems should
work in synergy and in a symbiotic relationship, since both are depending on each other and success
of AI systems lies in the mutual understanding of both.
Literature indicates that AI has many advantages over other technological innovations. Firstly, AI
can reduce the risk by supporting dynamic capabilities of sensing, seizing, and transforming
(Matilda and Chesbrough, 2020). Secondly, AI enlarges the scope of creative thinking (Eriksson et
al., 2020). Thirdly, some of the important characteristics supported by AI powered systems were
context-awareness, communication capability, embedded knowledge, reasoning capability and
self-organisation capability (Romero, Guédria, Panetto and Barafort, 2020). Fourthly, the
combination of AI, robotics and big data had been referred to as fourth industrial revolution due
the nature of immense impacts these technologies promise. Jarrahi (2018) had suggested AI
systems should not be designed with an intention of replacing human contribution but with the
intention of augmenting human knowledge and decision making.
The focus of this study is on exploring symbiotic relationship between employees and usage of AI
for making effective decision making in different elements of OM. The first research question
explored in this study is, how AI can be utilized in OM within an organization environment? For
exploring this research question, eight propositions (Proposition 1a, Proposition 1b, Proposition2a,
Proposition 2b, Proposition 3a, Proposition 3b, Proposition 4a and Proposition 4b) have been
developed. Proposition 1a and Proposition 1b explores the usage of AI within the manufacturing
element of OM for product inspection and quality function deployment. Proposition 2a and
Proposition 2b explores the usage of AI within the product development element of OM for
identifying core capabilities and self-learning products. Proposition 3a and Proposition 3b explores
the usage of AI within the services elements of OM for personalized targeting and enhancing
customer experiences. Proposition 4a and Proposition 4b explores the usage of AI within the supply
chain elements of OM for facilitating employee judgement and gathering customers need.
For exploring some of these propositions (Proposition 1a to Proposition 4b), two complementary
approaches, academic literature review and social media analytics, suggested by Grover, Kar and
Janssen (2019), for determining diffusion of blockchain in different industries had been used.
Literature review is a better approach for tracing features and challenges of the technology (in this
article technology under consideration is AI) within industry sector (in this article industry sector
under consideration is OM), whereas for tracing practical implementation of the use cases social
media is a better approach (Grover et al., 2019). Banomyong, Varadejsatitwong and Oloruntoba
(2019) had pointed out literature review identifies main research themes from the literature
available. Social media data had been used in academic literature for event classification (Singh et
al., 2019), electoral sentiment analysis (Grover et al., 2019a), reputation estimations (Grover et al.,
2019b), communication in emergency situations (Wamba et al., 2019), waste minimisation (Mishra
and Singh, 2018) and many more. The two sources of information provide a complementary
perspective for our analysis
The remaining sections are organized as follows. Section 2 is dedicated to background study, which
had been further divided into three sections OM, digital transformation and AI. Section 3 explains
theoretical background, research question and propositions explored in the article. Section 4
illustrates the research methodology adopted for the study. Section 5 presents finding from
academic literature review and social media analytics. Section 6 explains and illustrates discussions
of insights from academic literature and social media analytics. Subsequently, this is followed by a
conclusion section which discusses the limitations of the study along with future research
directions.
2. Background Study
This section contains three subsections, first subsection briefly illustrates OM definition and OM
evolution. The second subsection briefly introduces digital transformation and third subsection presents
AI and its tools and techniques.
2.1 Operations Management (OM)
Operation management (OM) has been segmented into three modules: “in the door”, “out of door” and
all the management activities between and beyond “in and out doors” (Karmarkar and Apte, 2007). The
first module, “in the doors”, takes care of the management activities required for getting appropriate
inputs. Sourcing, procurement, supplier selection and logistics are the major activities in this module. The
second module, “out of door”, takes care of the management activities required for sending goods and
services to the customers. This module focuses on three entities distributor, retailer and consumers
(Santiváñez and Melachrinoudis, 2020). The third module takes care of all the management activities
required between first and second module, requirement elicitation, production, co-creation and process
improvements within the firm.
Gunasekaran and Ngai (2012) had briefly illustrated the evolution of OM. In the beginning of the
evolution, the objective was on individual customer requirements whereby craftsman and artesian
production were the focused strategies. Immediately after the second World War, OM strategies shifted
to address objectives like demand for consumer products; total quality management, just-in-time and
transfer line production systems. In 1975 to 1985, OM strategies again shifted to focus on objectives like
addressing medium volume and variety; quick response manufacturing, computer integrated
manufacturing, flexible manufacturing systems and business process reengineering. In 1985 to 1995, OM
strategies shifted to focus on objectives like cost reduction, high variety and low volume; lean, agile and
physically distributed enterprise environments. In 1995 to 2010, OM objectives were shifted to higher
variety and very low volume; outsourcing, global manufacturing and market, agile, internet-enabled
supply chain management and third-party logistics. Further 2010 onwards, OM objectives shifted towards
global individualized products and services; total global supply chain management, virtual enterprise,
radio frequency identification enabled supply chain management, and sustainability. Gunasekaran and
Ngai (2012) had also pointed out how OM has evolved from mass production to mass customization
through these stages of evolution.
Subramanian and Ramanathan (2012) had categorized use of decision sciences in OM into five broad
themes: operation strategy, process and product design, planning and scheduling resources, project
management and managing the supply chain. In this article; manufacturing, product development,
services and the supply chain component of OM had been explored. Gunasekaran and Ngai (2012) had
illustrated that there is lack of quality standards as well there is an inconsistence in producing the quality
products across the globe. Morikawa (2017) had pointed out that both manufacturing and non-
manufacturing firms expect favourable impacts of AI on their businesses. Therefore in this study; through
proposition 1a, the possibility of using AI systems for product inspection and quality function deployment
in proposition 1b has been explored.
Subramanian and Ramanathan (2012) had defined product development as drawing up the specifications
for the making of the product appropriated to the customers’ needs. Proposition 2a explores the core
capabilities of the organizations on real time basis for product development through AI. Gunasekaran and
Ngai (2012) had pointed out that there is lack of customer knowledge on the product, therefore in
proposition 2b the usage of multi-agent distributed agents for self-learning products had been explored.
Further in proposition 3a and 3b, the use of AI for recommender systems for personalized targeting and
the use of intelligent chatbots (like Amazon’s Alexa or Apple’s Siri) for customer relationship management
has been discussed.
Supply chain encompasses all activities such as managing inventory, logistics, reverse logistics,
outsourcing (Queiroz and Wamba, 2019; Subramanian and Ramanathan, 2012). Supply chain can divided
into two parts upstream supply chain and downstream supply chain. Popular decision problems in
upstream supply chain is the selection of supplier (Kar, 2014; Kar, 2015). Gunasekaran and Ngai (2012)
had illustrated there is lack of fair practices in procurement and merit-based supplier selection. Therefore,
in proposition 4a, the usage of AI for facilitating human judgement had been explored. Subsequently in
proposition 4b, usage of AI for gathering customers need had been explored.
2.2 Digital Transformation
Romero, Guédria, Panetto and Barafort (2020) had characterised digital transformation as a change
of paradigm from computer-aided technologies to smart systems. ElMassah and Mohieldin (2020)
had highlighted digital transformation within an organisation is needed for decision making
purposes, whereas Bordeleau, Mosconi and de Santa-Eulalia (2020) had highlighted digital
transformation can leverage business intelligence and analytics within an organization.
Digital transformation had increased customer experiences and centricity (Taylor et al., 2019).
Digitization had evolved new marketing concepts such as value propositions and co-creation (Taylor
et al., 2019). Digitalization improves audit quality and firm governance (Manita et al., 2020).
Zangiacomi, Pessot, Fornasiero, Bertetti and Sacco (2020) had highlighted some of the practices for
digital transformation. Firstly, there is a need to understand how using digital technology company
business model can be changed (Job-fit). Secondly high level management should have an
awareness of the digital technology implications on the organization and employees (perceived
consequences and complexity). Berlak, Hafner and Kuppelwieser (2020) had highlighted
employee’s acceptance behaviour plays a significant role in generating productivity from digital
transformation (affect towards use). Behavioural intention has been pointed out as one of the
deciding factor of digital system use. Berlak, Hafner and Kuppelwieser (2020) had highlighted key
factors during digital transformation: attitude towards technology; perceived usefulness; desired
behaviour; and perceived ease of use impact this outcome more than anything else.
2.3 Artificial Intelligence (AI)
AI has been defined as a system’s ability to learn from external data correctly and apply the learnt
learnings for achieving specific goals and tasks (Haenlein and Kaplan, 2019). Such learning by the system
may be supervised, semi-supervised or unsupervised (Kar, 2016). Kumar, Rajan, Venkatesan and Lecinski
(2019) had defined AI as a tool for endless options and information that can be narrowed down to
personalized targeting (Kumar et al., 2019). Similarly Jarrahi (2018), had also pointed AI as a set of tools,
