**2. Literature review**

Based on the goal to measure areas of autonomy in warehouses, a literature review was performed. The following chapter will describe the methodological steps that were deployed to collect relevant information.

To begin with, relevant keywords for the search of the literature were derived. As the objective suggests, the main keywords would incorporate 'warehouse' and 'maturity'. Regarding warehousing in general, an important limitation was considered, namely the exclusion of 'data warehouses'. Therefore, the initial search term includes: 'warehouse maturity' – data warehouse. The search term is listed in quotation marks, which allows to search for exact matches in the title of a publication. Next, a search engine of google (scholar.google.com) was selected. This initial search for hits in the title led to only one result. Hence, the authors decided to enlarge the search to other functional areas including logistics (4.0), industry 4.0 and SCM in general. During


#### **Table 1.** *Selected studies from performed literature review).*

this first intervention, a higher amount of search hits was achieved. In total 11 additional hits were registered for the logistics 4.0 field. Next, 10 hits were counted for the area of industry 4.0 and finally, 5 relevant hits were registered for maturity models in SCM. For these results, titles, abstracts and summaries were analyzed. A second intervention was carried out and a deeper look at the publications revealed shortcomings, which led to the exclusion of 8 sources from logistics (4.0), 6 sources from industry 4.0 and 5 from SCM. On the other hand, a snowball-approach was deployed to investigate more into references made by relevant authors. This route led to 3 more interesting studies (2 from warehousing and 1 from industry 4.0). The final selection encompassed 11 papers that were further analyzed (**Table 1**).

#### **2.1 Maturity dimensions**

Since we were interested to find relevant dimensions in which the maturity of warehouses towards autonomy could be measured, we extended a request into our professional network of warehousing and SCM experts. The goal was to develop a reference process for warehousing that allowed to search for commonalities and deviations in the registered search results. These findings would eventually help us to confirm or to discard certain elements from the maturity model. The expert talks helped to confirm initial assumptions and extended the knowledge in this domain. As a summarization, the following essential process-steps were identified: deloading (of trucks), receiving, material handling and put-away, storage control (inventory management), picking, packing, loading (of trucks) and shipping. The expert rounds assisted in uncovering another selection of important functional areas that would complement the initial approach. These are the yard management as well as management of information technology, which is critical for the seamless flow of information, goods and finances. The dimensions identified were then mirrored against the papers to find commonalities and differences.

As shown in **Table 2**, the first comparison of the reference process with existing literature highlights important findings. In comparison to the Maturity Model of WERC [10], it is found that all aspects of the reference process, apart from de-loading and yard management, could be confirmed. Subsequently, the reference processes were investigated regarding the study of Salhieh and Alswaer. **Table 2** demonstrates that less commonalities could be identified. At least receiving, put-away, picking and shipping could be confirmed as overlapping process steps. Unlike the previous model, Salhieh and Alswaer clearly refer to maturity levels and recommend the usage of 5 different maturity levels. Lastly, the reference process was mirrored against the previous work of the authors. Related to an external project, the authors outlined a basic model to describe interactions and dependencies in modern warehousing. The comparison reveals that most of the process steps could be confirmed, apart from receiving, storage control, packing and shipping. As previously, 5 maturity levels were listed.

As outlined in the previous part of this chapter, the analysis was later extended into the area of logistics (4.0). This comparison is based on three studies. The first one covers the development of a maturity model for logistics 4.0 and includes a roadmap for further research [11]. The second paper goes into details regarding a framework for logistics maturity assessments, respectively with a portion that considers internal logistics [12]. Lastly, Asdecker and Felch present the development of maturity model for the delivery process in supply chains [13]. Unlike the previous comparison, the authors proceed without a detailed comparison against the reference process. The results were not fruitful enough and there were not enough commonalities to justify further discussion.

*Autonomous Warehousing: Development and Application of a Maturity Model DOI: http://dx.doi.org/10.5772/intechopen.104184*


#### **Table 2.**

*Analysis of commonalities and differences of warehouse reference process in warehousing literature.*

Contrasting, the section of literature related to industry 4.0 brought up interesting findings. Firstly, the reference process had to be adapted, since industry 4.0 covers more areas than warehousing alone. Commonly established models describe the vision of industry 4.0 with a variety of aspects: Business models, digital products & services, processes, production, social aspects, organizational factors, IT and digitalization and logistics. The first selected publication originates at the 'Competence centre for medium-sized business of North Rhine-Westphalia' and was later adapted by another company since public funding halted [14]. Next, the paper of Leyh et al. is referred [15]. The scientists consider assessing the IT and software landscapes of enterprises and propose a model for companies to increase their readiness for industry 4.0. Thirdly, the paper of Sony and Naik was selected, in which a literature review provides the basis to discuss the key ingredients for evaluating industry 4.0 readiness in organizations [16]. Fourthly, a study of Santos et al. is referenced, who discuss and propose an industry 4.0 readiness model [17]. Lastly, an additional study by Zoubek et al. was selected. It addresses a maturity model which could assist in evaluating and increasing the readiness within the concept of industry 4.0 while considering a focus on internal logistics processes (**Table 3**) [18].

Like the foregoing section, the goal of this comparison is to identify corresponding and deviating aspects regarding constituent elements of industry 4.0. Considering the information that can be extracted from the readiness assessment of NRW's competence centre, most of the elements seem to reappear. The same applies to the paper of Leyh et al., although the emphasis on digitalization, integration and cross-sectional technology implementation is stronger. Sony and Naik and Santos and Martinho mostly confirm previous findings but underline the importance of smart, respectively intelligent products, services and processes. From the comparison with the more general study of Zoubek and Poor, the emphasis is put on production and logistics as well as information technology.


#### **Table 3.**

*Analysis of design elements in industry 4.0 literature.*

## **2.2 Maturity levels**

Apart from the maturity dimensions that were discussed, a second approach was used to study the maturity levels of different models. While maturity dimensions point to the area where the measurement for maturity will take place, levels are used to assess a certain readiness in a particular area. To perform this assessment, it is important to consider that maturity levels should be chosen coherently and either in a quantitative or qualitative way. This definition also affects the interpretability of the results.

Regarding the first group of papers that were analyzed, mostly similar layouts of maturity levels were found. Due to a paywall, the maturity levels of WERC's maturity model could not be accessed. The research of Salhieh and Alswaer refers to 4 different maturity levels. To assess the maturity of areas like integrated warehouse performance measures, the two scientists propose levels starting from negligible, low, moderate or high. Each of the levels has more details to it, for example, a negligible maturity level would correlate with sub 25% usage or deployment of a certain measurement, while a high maturity level would correspond to the usage of performance measures that is in the range of 75–100%. Next, the investigations of Logistikum Schweiz GmbH resulted in 5 maturity levels. In the Warehouse Reference Process Model, various dimensions are addressed. As an example, the assessment in Yard Management Maturity refers to manual, mechanized, automated, digitally augmented or lastly, intelligent dark. To exemplify, the final intelligent dark maturity level would describe that all the work

#### *Autonomous Warehousing: Development and Application of a Maturity Model DOI: http://dx.doi.org/10.5772/intechopen.104184*

and services are done in an autonomous way, including autonomously operating robots. Regarding the findings in the industry 4.0 section, mostly consistent levels were found. The competence centre for medium-sized businesses suggests 5 different levels. The starting point is marked by paper transfer of data, transfer of paper data in digital form, general usage of ERP systems, and digital data completeness until the automatic transfer of data. In a similar fashion, Leyh et al. suggest the application of 5 maturity levels. To assess the maturity of the IT landscape the following levels are used: basic digitization, cross-department digitization, horizontal and vertical digitization, full digitization and optimized full digitization. While Sony and Naik would not address the details of maturity models, suitable considerations can be found for the studies of Santos and Martinho as well as for Zoubek and Poor. Both suggest 6 maturity levels in their models. Structurally, they only differ slightly from the previously investigated models. This differentiation is based on the first, initial maturity level, which is congruent for both, as their first levels start at zero actions, respectively zero shares of implemented initiatives.

To summarize this comparative representation, all reviewed maturity models relate to well-established components of industry 4.0. If this comparison is extended to warehousing, some notable differences come to attention. While it is obvious that industry 4.0 maturity models address basic functions like process management, production and logistics, additional elements like social, organizational and technological viewpoints are addressed as well. These elements are rarely represented in functional maturity models, like maturity models of logistics processes or applied technologies in logistics.

## **3. Model building approach**

Regarding the model development, a suitable methodological approach has to be chosen. Comparable investigations refer to the work of De Bruin et al. [9]. In their seminal paper, the scientists presented an often-cited approach that assists in developing specific maturity assessment models.

The approach is based on six subsequent phases: (1) scope, (2) design, (3) populate, (4) validate, (5) test and deploy and (6) maintain. Since this study is not a longitudinal study, only the first five phases will be used.

Phase 1: Scoping in the first phase of scoping, a decision must be made whether the model will address general or domain-specific use, which determines the scope and boundaries of the suggested model [9]. Apart from this decision, it is important to consider and include further stakeholders in the development of the model. This should ensure that possible benefits that result from the development or result from the use of the model can be shared with experts and vice-versa, experts can help and contribute to the model and its development stages. The exchange with science and industry is of great importance because it allows to build on existing knowledge and insights from previous research.

Phase 2: The design phase centres around five subsequent criteria, that determine the further layout of the model. They are intended to clarify the audience, the method of application, the driver of application, respondents and the application itself. For this study, the audience is mainly warehouse managers because they are directly involved in initiatives regarding the organizational and technological development of their facilities. Furthermore, the second audience of interest are consultants and auditors, who are often involved in guiding and accompanying warehouses. The principal

method of application will be mostly based on structured interviews. Based on the clarification of the reason, why such a model should be applied and for whom it will be developed, the next chapter addresses the remaining design aspects.

Phase 3: Populate: The next phase is centred around the population of the maturity models in terms of content and requires the description of model components and model subcomponents [9]. This description clarifies what content needs to be measured for any given component or subcomponent. According to DeBruin et al., various approaches are suited to define the contents of each subcomponent. For example, a thorough literature review could be suited, as well as empirical approaches such as stakeholder interviews, surveys, focus groups and in-depth case studies. For the present study, a combination of approaches was chosen. Firstly, a literature review was conducted to identify basic components and subcomponents. Secondly, the study used individual expert talks and semi-structured interviews to validate the findings and to check if certain aspects need to be further adjusted. The results of the literature review are discussed in a previous chapter. The feedbacks resulting from the expert talks largely confirmed initial viewpoints and assisted in validating the principal assumptions of the maturity model.

Phase 4: Model validating: The validity and reliability of the maturity model in scope are in the focus of the fourth phase of model development, according to DeBruin et al. validity and reliability are important building blocks to ensure and strengthen the relevance and rigor of the model [9]. While the validity of the model is supposed to secure the correlation between factual and intended measurements, the reliability addresses if the obtainable results are accurate and repeatable. As in previous sections, referring to DeBruin et al. reveal different approaches to ensure such requirements. Surveys, interviews or literature reviews are among the options to be used in this regard. For the present study, all maturity dimensions and maturity levels are grounded in previous research publications and were validated in expert interviews. Therefore, the validity and reliability are confirmed.

Phase 5: Test and deployment. Within the last consecutive phase of model development guidelines [9], the deployment of the model is addressed. Following the guidelines of DeBruin et al., this phase aims to clarify the generalizability of the model pursued. This can be achieved by applying the model within suitable case studies. DeBruin et al. refer to two separate approaches. In the first place, it is suggested to test the model within an audience consisting of stakeholders, who were directly or indirectly involved in the development of the model itself. Secondly, the generalizability can be extended by discussing the model with an audience that is not part of a stakeholder group, respectively is external to the domain of warehouse maturity research. By pursuing these two steps, the general acceptance of the model can be reflected and confirmed. For the research at hand, the testing of the model was done by discussing it with members of a specific focus group that is committed to develop the warehouse of the future and consists of various warehouse operators, consultants and solutions providers with extensive experience in this domain. As a result of discussions, it could be found that the majority agrees with the proposals made by the authors. These discussions not only helped to ensure acceptance and demand for the model but also assisted in enriching the initial building. The focus group mentioned above also involves research partners from two universities of applied sciences. From each of the partners, one person who was not involved in the model development was asked to review the model and provide feedback to the authors. Like the first group, the authors could not learn about contributions that would question the current state of the model. To summarize, this initiative led to further confirmation of the model proposed.

*Autonomous Warehousing: Development and Application of a Maturity Model DOI: http://dx.doi.org/10.5772/intechopen.104184*
