Demystifying Industrial Internet of Things start-ups – A multi-layer taxonomy
Abstract:
Described as a fundamental paradigm shift by researchers, the Industrial Internet of Things (IIoT) is credited with massive potential. In the context of emerging technologies, such as the IIoT, start-ups occupy a crucial role, as new technologies are often first commercialized by start-ups. Because of the rising importance of IIoT start-ups as drivers of industrial innovation, IIoT solutions demand deepened theoretical insights. As existing classification schemes in the industrial context do not sufficiently account for the ever more critical role of IIoT start-ups, we present a multi-layer taxonomy of IIoT start-up solutions. Building on state-of-the-art literature and a sample of 78 real-world IIoT start-up solutions, the taxonomy comprises ten dimensions and related characteristics structured along the three layers solution, data, and business model. The taxonomy contributes to the descriptive knowledge on the IIoT and enables researchers and practitioners to better understand IIoT start-up solutions.
Keywords: Industrial Internet of Things, Industry 4.0, Start-up, Solutions, Taxonomy 1
Introduction:
Without a doubt, one can state: The Industrial Internet of Things (IIoT) is among the most discussed industrial business concepts in recent years and is seen as a fundamental paradigm shift in industrial production [2, 3]. Experts are already forecasting a market size of USD 110.6 billion for the IIoT in 2025 [5]. The number of connected devices is expected to exceed the magic mark of 50 billion by 2030, highlighting the potential of the technology [6]. The IIoT refers to the extension and use of the Internet of Things (IoT) in manufacturing, enabling industrial systems’ interconnection to improve productivity, efficiency, safety, and intelligence [8]. In the context of emerging technologies, such as the IIoT, start-ups occupy a crucial role, as new technologies are often first commercialized by start-ups and, therefore, are drivers for innovation [9]. Because of that, IIoT start-ups are indispensable partners in the digital transformation of incumbent industrial companies, as they often supply the innovative IIoT solutions needed. In this paper, we understand IIoT start-ups as newly established businesses that offer IIoT solutions for the business-to-business (B2B) market. Examining not IIoT specifically, but IoT in general, IoT start-up venture capital saw a 15% year-on-year increase from Q1 2019 to Q1 2020, with a total of USD 4.7 billion [10]. CrunchBase, an investment platform for start-ups, shows a 27% increase in IoT start-ups and related businesses from 26.792 to 34.120 (as of May 2020) in just one year [11]. Within the IoT, the IIoT is attributed with considerable economic potential [12–15]. The impressive numbers of start-ups in the IIoT sector reinforce the need to analyze IIoT start-up solutions in research. One example of such an IIoT start-up solution is provided by Aspinity, which patented a unique modular processor technology, enabling a system-level solution that overcomes the power challenge for always-on edge processing [16]. Moreover, TeleSense provides remote solutions for grain storage and transportation monitoring [17]. On the one hand, these two examples show that IIoT startups are forerunners in developing and implementing new IIoT solutions. On the other hand, it shows the variety of solutions offered by IIoT startups, ranging from monitoring solutions to more complex processor technology [9]. Despite the increasing relevance of IIoT as a paradigm shifter, little theoretical insights exist about the companies which often first commercialize IIoT solutions, namely start-ups [9]. Existing IIoT classifications focus on topics such as business models [18], platform features [19], and industrial service systems enabled by digital product innovation [7]. Hence, creating a deepened theoretical understanding of IIoT start-up solutions that drive the global manufacturing paradigm shift is useful and valuable.
(Industrial) Internet of Things:
The term “Internet of Things” was first coined in 1999 to describe uniquely identifiable, interoperable, connected objects using radio frequency identification technology in the supply chain [22]. Oberländer et al. [23] define IoT as “the connectivity of physical objects equipped with sensors and actuators to the Internet via data communication technology” (p. 489). Using these sensors and actuators, it becomes possible to connect the physical world to the Internet [24]. These technology-embedded objects are also called “smart things” [1]. IoT can be categorized into three domains, namely Consumer, Commercial, and Industrial [15]. Consumer IoT addresses the business-to-consumer market and refers to use cases such as smart refrigerators or smart thermostats [1]. Looking at the IoT B2B market, a distinction between Commercial and Industrial IoT can be made [15]. Use-cases such as connected medical devices or inventory controls refer to the Commercial IoT [15]. In contrast, the Industrial IoT connects industrial devices, production facility systems, and manufacturing processes [25, 26]. A widely used definition of the term IIoT is provided by Boyes et al. [27]: “A system comprising networked smart objects, cyber-physical assets, associated generic information technologies and optional cloud or edge computing platforms, which enable real-time, intelligent, and autonomous access, collection, analysis, communications, and exchange of process, product and/or service information, within the industrial environment, so as to optimize overall production value” (pp. 3-4). Thus, IIoT – or Industry 4.0, as it is known in the German-speaking community [25, 28] – refers to the application of IoT in industrial manufacturing and has the potential to improve productivity, efficiency, safety, and intelligence of industrial operations [3, 8, 29]. The improvement is made possible by the interconnection of different industrial systems (machines, control systems, and information systems) and collected data, enabling analytic solutions leading to optimized industrial processes [30]. As IoT and IIoT rise in popularity and importance globally, start-ups actively participate in the IoT and IIoT industry [31]. Start-ups represent fast-growing ventures that serve a need in the marketplace by offering innovative solutions [31, 32]. As indicated above, IIoT holds considerable market potential, also reflected in start-up funding [10]. Since there is no commonly agreed-on definition of IIoT start-ups, we define IIoT start-ups in the context of this paper as a composite of the two terms IIoT and start-up: IIoT start-ups are newly established businesses that offer IIoT solutions for the B2B market. An IIoT solution is an integrated offering that can be either a product, a service, or both to create a smart industrial environment that delivers value for the B2B customer [33, 34].
Related Work Informing the Taxonomy of IIoT Start-up Solutions:
Taxonomies help understand and analyze complex domains by grouping objects based on common characteristics and analyzing the relationships between the taxonomy’s characteristics [20, 35, 36]. Thereby, the terms for structuring concepts – taxonomy, framework, or typology – are used as synonyms [4, 37]. Taxonomies are especially suitable when little knowledge exists [21]. A taxonomy is excellent for analyzing the multitude of IIoT start-up solutions since IIoT start-ups are an emerging phenomenon and, thus, little theoretical understanding exists. Below, we briefly introduce relevant existing taxonomies for our purpose. In terms of taxonomy design, we were able to incorporate several things from existing taxonomies. First, a second-level grouping allows for better comprehensibility of a taxonomy; for instance, Gimpel et al. [37] use a second-level grouping to classify FinTech start-ups’ service offerings. Second, we learned that some taxonomies include non-exclusive dimensions. For example, Püschel et al. [1] present a taxonomy with non-exclusive dimensions to understand the non-technical characteristics of smart things to tap the full potential of smart things. Concerning the layers, dimensions, and characteristics of the selected taxonomies, we were able to identify the following aspects. Since most IIoT start-ups consider data a critical resource for their business operation, the taxonomy of Hartmann et al. [38] on data-driven business models of start-ups was particularly relevant in terms of data sources. As IIoT start-ups operate in an industrial context, the taxonomy of industrial service systems by Herterich et al. [7] was suitable, as we could generate further insights about the relevance of data from an industrial perspective. Furthermore, Rizk et al. [39] take a data analytics perspective to classify data-driven digital services, which was helpful because analyzing data is an essential part of IIoT start-up solutions. Paukstadt et al. [4] provide a taxonomy to classify smart services along the three layers service concept, service delivery, and service monetization, focusing on the specific characteristics of smart services. Thereby, smart services are defined as services enabled by smart products [4, 40]. The taxonomy by Paukstadt et al. [4] was especially relevant, as it fosters an understanding of possible descriptions and forms of smart services. Since smart things are also part of the range of solutions provided by IIoT start-ups, relevant dimensions could be obtained from the taxonomy about smart things of Püschel et al. [1]. We used the existing taxonomies as a starting point to develop a new taxonomy, enabling the classification of IIoT start-up solutions.
Research Method :
To answer our research question and address our target users, such as IIoT researchers and practitioners, we opted for a rigorous method approach already well established in the literature for taxonomies by Nickerson et al. [20]. The iteration-based approach by Nickerson et al. [20] combines qualitative and quantitative research [41]. Thus, it is allowed to use both academic literature and empirical objects to develop layers, dimensions, and characteristics. Furthermore, this iterative approach is widely used to structure complex and emerging fields where little knowledge exists [21, 42]. A taxonomy describes a classification strategy for grouping objects [20]. Before starting with the individual iterations, the meta-characteristic and the corresponding objective and subjective ending conditions have to be set, which serve as the foundation of the taxonomy and describe when the iterative process can be terminated [20]. For each iteration, the conceptual-to-empirical or the empirical-to-conceptual approach can be chosen [20]. In a conceptual-to-empirical iteration, the layers, dimensions, and characteristics are based on the literature or the author’s knowledge. In an empiricalto-conceptual approach, a sample of real-world examples gets analyzed. After finishing an iteration, an initial or revised taxonomy is obtained, and the authors must check whether the ending conditions are met. The taxonomy development process continues until the objective and subjective ending conditions are met. The purpose of our taxonomy is to enable researchers and practitioners to understand and classify the diverse solutions offered by IIoT start-ups. To start the taxonomy development process, we first defined the meta-characteristic of our taxonomy. In line with our research question, our meta-characteristic was “classification of IIoT start-up solutions offered in the context of B2B”. Second, we decided on the ending conditions. By choosing from a list of proposed objective ending conditions by Nickerson et al. [20], we came up with the following objective ending conditions: (1) each characteristic is unique within its dimension, (2) each dimension is unique and not repeated within the taxonomy, and (3) at least one object is classified per characteristic and dimension. Further, we chose five subjective ending conditions, which are met if the authors agree that the taxonomy is concise, robust, comprehensive, extendible, and explanatory [20]. Additionally, Nickerson et al. [20] require characteristics to be mutually exclusive. However, it is not possible for some dimensions to restrict the choice of characteristics to be mutually exclusive, as relevant information would be lost. In line with other published taxonomies, e.g., Gimpel et al. [37] and Püschel et al. [1], we allowed nonexclusive dimensions. Our taxonomy development process comprised four iterations. 1 Iteration: In the first iteration, we opted for the conceptual-to-empirical approach, as IIoT start-up solutions comprise a relatively young and dynamic field of research. We conducted a short literature review to accumulate sufficient information about taxonomies related to IIoT start-ups solutions and adjacent or overlapping research fields (see Section 2.2). In line with the proposed meta-characteristic, we extracted initial dimensions and related characteristics to capture the first distinct features of IIoT start-up solutions. The conceptual-to-empirical approach led to a rudimentary taxonomy and built the foundation for the upcoming iterations. Since the rudimentary taxonomy depicted the taxonomy’s characteristics at different granular levels, the subjective ending condition “concise” was not met. Hence, a second iteration was conducted. 2 Iteration: We enhanced and validated our taxonomy’s structure by applying the empirical-toconceptual approach. To find real-world objects, we relied on CrunchBase, which claims to be the primary source of start-up insights listing over one million start-ups [43]. As part of generating a randomized sample from CrunchBase, we used several keywords (“IIoT” AND “Industrial Internet of Things” AND “Industry 4.0”) in our search string. This approach led to a representative sample size of 90 randomly drawn IIoT start-ups from a total of 626 hits. However, to guarantee comparability among IIoT start-ups and their respective solutions, we reduced the number of suitable IIoT startups according to the following criteria: (1) the CrunchBase website or the IIoT start-up website must provide sufficient information, and (2) the IIoT start-up must comply with our definition of IIoT start-ups: IIoT start-ups are newly established businesses that offer IIoT solutions for the B2B market.
Conclusion:
Despite the increasing relevance of IIoT as paradigm shifter, little insight exists about the companies which often first commercialize IIoT solutions, namely start-ups [9]. IIoT start-ups are indispensable partners in the digital transformation of incumbent industrial companies, as they offer a wide variety of IIoT solutions. To answer the research question of how IIoT start-up solutions can be classified, we proposed a multi-layer taxonomy that follows Nickerson et al.’s [20] established taxonomy development process. First, we reviewed existing literature to identify relevant dimensions and characteristics of IIoT start-up solutions [1, 4, 7, 37–39]. We then analyzed a randomized sample of 78 IIoT start-ups solutions from CrunchBase in four iterations until the objective and subjective ending conditions were met. From a theoretical perspective, our taxonomy contributes to the descriptive knowledge on the IIoT start-up phenomenon, exploring a not yet well-understood research field. Our main contribution is a theoretically well-founded and empirically validated taxonomy. The taxonomy serves as a starting point for researchers for further theorizing, e.g., for deriving archetypes (e.g., [18]) and theories for analyzing or explaining. On the one hand, archetypes help to understand higher-order configurations of IIoT start-up solutions and to anticipate trends within IIoT and related industries. On the other hand, the taxonomy constitutes a building block for developing a theory for analyzing IIoT startup solutions, e.g., by describing the phenomena, relationships, and boundaries [21]. From a practical perspective, our taxonomy serves as a tool for various players within the field of IIoT. Our taxonomy provides transparency from the perspective of an industrial company looking for a partner to implement an IIoT initiative. Our taxonomy enables the analysis of the various solutions offered by IIoT start-ups, e.g., how many IIoT start-up solutions are third-party integrable. From the viewpoint of an IIoT start-up, our taxonomy could serve as a basis for creating a market overview, finding niches, and examining them for their respective market potential, e.g., our taxonomy shows that certain areas are hardly addressed within the field of IIoT. In addition, our taxonomy helps to understand the phenomenon of IIoT start-ups better, identify core solutions, and define typical solution characteristics. Although this paper provides initial theoretical and practical implications, our study has its limitations and, thus, stimulates further research. First, our sample of IIoT startups is not exhaustive, as we only classified 78 randomly drawn IIoT start-ups. Future research should analyze more IIoT start-ups from different databases. Second, in some cases, the pricing information was non-transparent, possibly causing characteristics frequencies to be even higher or lower than observed. Third, the field of IIoT start-up solutions is dynamic. Therefore, our taxonomy represents a snapshot, as emerging types of IIoT start-up solutions may be underrepresented. Re-evaluating the dimensions and characteristics after a certain period is recommended, as this will provide longitudinal insights regarding the development of IIoT start-up solutions. To address the limitations above in further research, we developed the taxonomy as Nickerson et al. [20] suggested. Hence, the taxonomy is revisable and expandable. Further, we believe that this paper is of theoretical and practical relevancy. Thus, we hope to inspire fellow researchers to continue the research on IIoT solutions in the context of start-ups.
Download the Link Here :