CAPABILITIES FOR D TIES FOR DATA ANALYTICS IN INDUSTRI TICS IN INDUSTRIAL INTERNET OF THINGS (IIOT)

Abstract:

The use of Industrial Internet of Things (IIoT) technologies in various industrial sectors has resulted in the generation of large volumes of data that can be analyzed using analytics tools to improve firm performance. However, there is a gap in our understanding of the capabilities that companies need to create business value through data analytics in IIoT environments. Although previous research has extensively investigated general data analytics capabilities, the literature on these capabilities cannot be simply transferred to IIoT settings due to the unique characteristics of the IIoT. In this paper, we aim to contribute to our understanding of this phenomenon by identifying the capabilities required for IIoT data analytics. Firstly, we identify data analytics capabilities from existing literature. Next, we investigate the relevance of these capabilities in the context of IIoT, while also identifying novel capabilities that are specific to IIoT, by conducting 16 expert interviews within nine organizations. We identify a set of 24 capabilities for data analytics in IIoT, which we classify into an integrative framework. The proposed framework can assist industrial companies dealing with the complexities and uncertainties associated with IIoT data analytics initiatives.

Keywords: Industrial Internet of Things, IIoT capabilities, data analytics capabilities, Business value.

introduction:

Today Internet of Things (IoT) technologies are drawing considerable attention from both academics and practitioners. Many organizations have been heavily investing in these technologies (Siow et al., 2018). McKinsey estimates that the number of businesses that use IoT technologies has increased from 13 percent in 2014 to about 25 percent in 2019 (Dahlqvist et al., 2019), and it predicts that IoT could enable $5.5 trillion to $12.6 trillion in value globally, by 2030 (Chui et al., 2021). Industrial Internet of Things (IIoT) as a subset of IoT rapidly transforming the landscape of many industrial sectors, (e.g., smart manufacturing, logistics, retail, and utilities) by enabling smart automation, intelligent decision-making, and data-driven optimization (Boyes et al., 2018, Khan et al., 2020, Qi et al., 2023, Rehman et al., 2019). IIoT refers to the network of intelligent and highly connected industrial components that are deployed to achieve a high production rate with reduced operational costs through real-time monitoring, efficient management, and controlling of industrial processes, assets, and operational time (Khan et al., 2020). In mission-critical industrial settings, IIoT requires higher levels of safety, security, and reliable communication without the disruption of real-time industrial operations (Qi et al., 2023). However, IIoT devices such as sensors, RFIDs, and actuators integrated with industrial equipment generate large volumes of diverse and heterogeneous data, which can be challenging to process due to restricted computing, networking, and storage resources (Boyes et al., 2018, Qi et al., 2023).

To derive business value and gain a competitive advantage, IIoT data must be collected, integrated, and analyzed in a scalable and cost-effective manner (Ahmed et al., 2017). To achieve this, data analytics tools play a vital role in transforming IIoT data into valuable business insights (Ahmed et al., 2017, Boyes et al., 2018, Côrte-Real et al., 2020, Rehman et al., 2019). However, incorporating data analytics in the context of IIoT presents various challenges, including interoperability and system integration difficulties, data offloading, missing data streams, network performance issues (such as congestion, delays, and bandwidth), system architecture, scalability, reliability, and security and privacy concerns (Khan et al., 2020, Qi et al., 2023). Nevertheless, in prior studies, most emphasis has been on the technical aspects of data analytics in IIoT environments (Ahmed et al., 2017, Côrte-Real et al., 2020, Marjani et al., 2017) while limited attention has been paid to understanding how to leverage them to generate business value. Previous research on business value creation through data analytics has shown that while data analytics tools are necessary, they alone are not enough to produce business value (Gupta and George, 2016, Mikalef et al., 2018). 

One lens to study how firms leverage data analytics to realize performance gains is data analytics capabilities (Gupta and George, 2016, Mikalef et al., 2018, Wang et al., 2018). Drawing on the resource-based theory, data analytics capabilities refer to a company’s ability to effectively collect, analyze, and interpret data by effectively managing and utilizing its data, technology, and talent to generate business insights (Akter et al., 2016, Gupta and George, 2016, Mikalef et al., 2018). While data analytics capabilities have been widely investigated in prior research, such as (Akter et al., 2016, Mikalef et al., 2019, Wang et al., 2018), due to idiosyncrasies of the IIoT data (like heterogeneity of data sources, real-time streaming data, data redundancy, limited communication range, restricted networking, and node failure (Boyes et al., 2018, Côrte-Real et al., 2020, Qi et al., 2023), the literature on data analytics capabilities cannot be simply transferred to the IIoT settings. As with any new technology, such as that of IIoT, organizations need to develop a unique set of capabilities to effectively leverage their investments to generate business value (Mikalef and Gupta, 2021). Identifying the necessary capabilities for leveraging data analytics in IIoT settings is crucial to ensure its benefits are realized. 

This paper aims to identify and analyze the capabilities required for the successful implementation of data analytics in the Industrial Internet of Things (IIoT) setting to create business value. Specifically, the paper addresses the following research question: Which capabilities are required for data analytics in IIoT environments to generate business value? We approach this question by, first, identifying capabilities from the data analytics literature. Second, we investigate if and how these capabilities are relevant in the context of IIoT by conducting 16 expert interviews across 9 organizations. The outline of the paper is as follows. Related work is discussed in Section 2. Section 3 describes the research methodology. The results of the expert interviews that led to the proposed framework of capabilitiesfor data analytics in IIoT are presented and discussed in Section 4. Finally, Section 5 presents the conclusion. 

Capabilities and Business Value of Data Analytics in the IIoT:

The identified 24 capabilities for data analytics in the IIoT environments have the potential to generate two types of business value: strategic and operational. As IIoT is integrated into vital business processes, the operational business value tends to take precedence. The participants in the study emphasized that organizations are primarily interested in investing in analytics to gain real-time insights through visualizations and dashboards, and subsequently to achieve operational excellence via advanced use of IoT data and automation. Drawing from both the review of data analytics capabilities literature and the empirical data analysis in the IIoT context, a unified framework that encompasses capabilities for data analytics in the IIoT and the resulting business value is presented in Figure 1.

Conclusion:

By investigating the differences between generic data analytics (Table 1) and capabilities required for the successful implementation of data analytics in the IIoT setting (Tables 3-5), this study shows that generic data analytics capabilities can largely be transferred to the context of IIoT. One major observation, however, is that data analytics in the context of IIoT is challenging because of the novel and often complex characteristics of IIoT that can give rise to specific data-related issues. For instance, IIoT systems require real-time data processing capabilities to ensure that decisions are made quickly and accurately. This can be a challenge when dealing with large volumes of data and when there are strict latency requirements. In this study, we identify the capabilities required for IIoT data analytics by conducting 16 expert interviews from nine big industrial companies operating in the logistics, smart manufacturing, and utility sectors, that use IIoT technologies and implement data analytics within the IIoT context. We identify a set of 24 capabilities including ten unique IIoT capabilities: IIoT platform architecture & design, Edge & hardware development, Software development, Connectivity, IIoT data processing & standardization, Operational maintenance & monitoring, Data accessibility, Knowledge management & training, Business and ecosystem synergy, Product and service development. Our work has a two-fold contribution to the IIoT literature. Firstly, we enrich the existing knowledge of data analytics in the IIoT context by identifying a set of capabilities that are imperative for the implementation of data analytics. We also analyze how these capabilities impact firm performance, resulting in a practical yet theoretically grounded framework of integrative capabilities and business values of data analytics in IIoT. This framework comprises 24 capabilities and encompasses both strategic and operational business values. Secondly, we also identify and discuss 10 capabilities that are specific to IIoT, highlighting the unique requirements for data analytics in this domain that differ from those of generic data analytics capabilities. By distinguishing these capabilities, we provide a more nuanced understanding of the distinctive needs of data analytics in the context of IIoT. Practitioners can benefit from this study in three ways. Firstly, the study highlights the differences between generic data analytics and analytics in the context of IIoT. This indicates that practitioners need to deal with unprecedented levels of complexity and scale, as well as gain new and specific knowledge about potentially unique technologies in the IIoT context. Secondly, the proposed framework can assist organizations in dealing with the uncertainties and complexities often involved in IIoT and analytics projects, enabling them to develop a comprehensive plan and identify the appropriate business goals for analytics initiatives in the IIoT context. It is crucial for organizations to understand that, in addition to generic data analytics capabilities, a range of technical and organizational capabilities specific to IIoT is required for the successful implementation of data analytics initiatives. Thirdly, the skills and knowledge required to implement data analytics in the IIoT context should not be underestimated. Furthermore, managers must realize that data analytics in the IIoT context necessitates a long-term and pragmatic vision, which can facilitate the organization's transition towards a more integrated future. This study is constrained by a number of limitations that open an avenue for future research. First, while our attempt is to provide a structured overview of the literature on data analytics capabilities, due to potential bias in the coverage of the literature, we do not claim that a list of data analytics capabilities is complete. Future research may expand on this list. Second, we interviewed 16 experts from nine big industrial companies. Although our purposive sampling approach involved a cross-organizational and cross-industry sample, we acknowledge that it may not have fully captured the potential specificities of companies or industries. Nevertheless, this strategy helped us minimize bias and achieve a certain level of confidence in the generalizability of our findings to companies operating in logistics, smart manufacturing, and utility sectors and implementing data analytics in the context of IIoT. However, the generalizability of our findings to other industries may require further research. Third, this research is a first step towards understanding business value creation through data analytics in the context of IIoT from the theoretical lens of capabilities. Therefore, the contribution of this study is largely explorative. Our claim is not that the proposed framework is complete but rather that it provides a basis for further empirical investigation and validation. Fourth, specific IIoT capabilities from our expert interviews may be subject to the personal bias of the interviewees. They might have a different understanding of capabilities or do not have complete knowledge of the capabilities required for data analytics in IIoT settings. We undertook considerable efforts to at least partially mitigated this problem, for instance, by providing interviewees with the definition of each data analytics capability (from table 1) as well as interviewing multiple experts from the same industries. Interviewing multiple respondents within a single organization would be useful to improve internal validity. Finally, the proposed framework can serve as a valuable starting point for understanding capabilities required for analytics in the IIoT context. Nevertheless, our exploratory study did not examine the impact of each capability on strategic and operational business value. Future explanatory studies should investigate the interrelationships and linkages between these capabilities and business value. 

                                Download the Link Here :