Evaluation of the Economic, Ecological and Ethical Potential of Big Data Solutions for a Digital Utopia in Logistics

Abstract

In the context of digital transformation and the use of big data technologies, this study shows the potential and possible applications associated with using big data, depending on the respective logistics area. The evaluation of this potential follows a logistical target system, which has been expanded to include ethical and sustainable aspects in line with the challenges of the 21st century. Furthermore, the status quo of the degree of digitization is determined, problems in understanding the terminology are shown, and cognitive and technical prerequisites and recommendations for action concerning the use of big data technologies (e.g., cobots) are derived. The survey methodology was based on a quantitative research method in the form of a survey. The article aims to provide the building blocks for a holistic approach (economic, ecological, and social) for politics and companies and to derive recommendations for action in logistics. The challenge is to find an alternative to conventional research, which examines big data almost exclusively for growth targets and optimization potential. Ecological and social goals should also be included here as an unchangeably fixed point and a normative limit as a value compass for business decisions. To close this research gap, it is necessary to evaluate the potential of logistics in connection with big data solutions to derive specific applications, their applications, and recommendations for action. This article presents an excerpt of the results.

1. Introduction

In the business administration of the early 20th century, besides the knowledge of antiquity, the writings of the fleet admiral Cyrus Thorpe had a decisive influence on the concept of logistics. By using historical examples, Thorpe explains in his book Pure Logistics in 1923 the necessity of good logistics and the maintenance of logistical skills such as planning and organization [1]. At the end of the 19th century, amid the first two industrial revolutions, the concept of economic efficiency was raised to a new level. The young engineer Frederick W. Taylor was particularly interested in manufacturing processes and, above all, the related question of how to organize and design them more efficiently. Skilled workers were considered expensive and were avoided to lower the cost. It was not the person but the system that came first. “In the past, humans came first; in the future, the system must come first. The first objective of any good system must be the development of first-class people” [2]. Taylor consistently applied the results of the methods from the work processes analysis.

The logistics of the 21st century can no longer focus solely on purely economic objectives, but must face new challenges, particularly of an ecological and social nature. The climate crisis [3], global resource consumption, which has long exceeded the limits of resilience [4], supply chains that start with inhumane working conditions [5], or costs that are not borne by the polluter but by society [6] are just a few examples of the precarious situations the world finds itself in. The effects on the climate and the environment, which emanate from CO₂ emissions, for example, have been published numerous times. Political goals for climate protection and specifically for limiting global temperature rise have already been defined in the 2015 Paris Climate Agreement [7].

There is an urgent need to identify the causes of these crises and systematically develop solution concepts that must be adapted and implemented for the area of responsibility in logistics. In this way, new goals in logistics can and must be derived. In addition to economic goals, they should also address sustainability aspects and social issues. Due to increasing digitization, which has reached the logistics industry to a greater or lesser extent, associated opportunities and risks are directly related to mastering these significant tasks. Based on these challenges, several theses could be derived. In essence, the thesis describes the potential of digital technologies, the prerequisites for employees’ specific knowledge, and critical technical prerequisites.

The following research questions were established and answered in the form of partial results:

(1)

What are the potentials and areas of application in the sub-areas of logistics that can be identified through big data technologies?

(2)

What are the fundamental problem areas identified through the use of big data technologies in logistic companies?

(3)

What potential added value could be achieved in the individual areas of logistics with the help of big data technologies?

(4)

Which meaningful structure of the contact groups could be determined so that a holistic approach can be pursued?

All these scientific questions are based on the approach of an economic, ecological, and social perspective. This is an approach based on the triumvirate of sustainability and, at the same time, serves as the basis for sustainable management oriented towards the common good [8].

In preparation for the writing of this article, two subject areas were evaluated to identify the potential. Firstly, logistics were examined regarding their structure and nature, and secondly, the ‘Big Data’ topic included all relevant sub-topics. If these two subject areas are brought together, this can be regarded as a new scientific approach, resulting in the new logistics objective mentioned earlier.

1.1. New Target System for Logistics

With digital transformation, big data technologies have also found their general place in everyday business. However, these technologies already serve a wide range of applications in the financial sector and are primarily understood as a black box in logistics. There is controversy about what potential benefits and dangers arise with the increasing digitization, which is almost complete in Central Europe. At the societal level, new technologies are changing public life at an unprecedented rate. “Big data: the greater good or invasion of privacy?” [9]. This quote from 2013 shows that digitization reaches all areas of the economy and society.

The goal is now to break this development down to the logistics industry and to describe a future scenario that considers economic, ecological, and social goals. It is necessary to identify various factors that influence decisions in logistics driven by big data technologies to achieve the plans according to the logistics target system (in Anlehnung an Wiendahl und Wiendahl) [10]. Figure 1 illustrates a logistics’ comprehensive target system.

With a historical view of logistics, general goals can now be defined against which future logistics (Logistics 4.0) must be measured [11]. Subsequently, it is necessary to harmonize the economic, ethical, and ecological potentials contained in the goals. Likewise, one must remember urgent conflict management, as conflicts sometimes arise between goals.

The overarching goals of logistics are primarily related to reducing the cost of logistical activities and improving the value and utilization of physical goods and services. The focus is also on enhancing the flexibility of logistic systems and adapting to constantly changing environmental conditions and influencing factors. The importance of logistics in a company has been changing continuously over the past decades and has grown steadily. The main reasons for this are the globalization of the markets and value chains, which are spread across the globe, especially in the case of global players. The increasing understanding of logistical and extensive process chains with information and communication technologies plays a role here, as does the desire for individualized products in the consumer goods markets and the supply-oriented economic policy of the transport and telecommunications markets [12]. However, this steady economic growth is often contrary to an ecologically and ethically justifiable economy. More growth usually means more consumption of resources [13].

Efficiency can therefore be viewed as logistics’ general and initially overarching economic goal. The costs of the required logistical processes performance should be kept as low as possible to guarantee simultaneous maximal and optimal performance. The holistic and customer-oriented view represents an essential basis. Overall, the total costs that arise in logistics must be considered. This includes the logistics chain and the full service for the customer. In the meantime, ecological and ethical goals and sustainability are becoming increasingly important in the logistical process. Whether and how this thought finds its way into reality, and therefore into society’s economic and ecological awareness, will decisively and sustainably shape the future of coexistence.

Viewed through the lens of economic glasses, the logistics target parameters can be divided into two main objectives. Firstly, the primary goal is a high performance, from which a simultaneous service of several customers’ results and a higher turnover can be generated as a direct consequence. Secondly, minimal logistics costs for transport, freight, storage, processing, or customer service fees should be incurred [10].

The logistics of the 21st century will have to face new challenges than just continuing the essentially pure efficiency concept of the 20th century and optimizing processes in terms of their efficiency and costs. The central question of the 21st century will be one of an ecological and social nature. To adequately answer these questions and thus meet the challenges of the 21st century, value systems must be rethought simultaneously, followed by more social and sustainable action. Its basis should be a newly designed framework for a successful Logistics 4.0 and the continuation of a prosperous economy that is in harmony with nature and society. Thus, the conventional target system of logistics must be expanded by four additional framework-giving target values (See Figure 1).

1.2. Understanding the Topic

From an economic point of view, the identification of potential is paramount. The reason is that these potentials still need to be sufficiently collected, and if adequately researched, they form the basis for recommendations for action. However, the challenges of the 21st century require a holistic approach that focuses on more than the economic aspects. Questions were also raised about ecological and ethical logistics, the structure of which is based on the new target logistics system. The survey results should serve as a basis for developing a value concept, which conveys starting points and recommendations for action that promote the common good and, at the same time, points out factors that are worthy of criticism in the system established today. Furthermore, factors influencing ideas and AI-driven logistics decisions should be identified. In this context, the consistent recognition of dangers through big data technologies and the development of recommendations for action that implement an integrative vision is part of further research. Well-founded basic research, which includes a precise definition of terms, is necessary for these goals.

1.2.1. Definition of Big Data

Big data, a term enjoying its increasing popularity since 2011/2012, is spreading exponentially. Unlike in the purely defined realm of computer science, big data is on everyone’s lips, even in the media and the public sphere. However, the rapid development of the digital economy and the impact of available technologies also leads to a compulsion to use devices that can process large amounts of data, which in turn should improve the performance of these companies, but there is also distrust and fear towards them that this new technology entails [14]. There is talk of active media work with data, of large amounts of data: big data [15]. The term big data as such is not clearly defined, nor is its authorship [16]. Large amounts of data are associated with data storage and subsequent data analysis. The definition that big data is only about storing large amounts of data is only partially sufficient for the description and only partially exploits the potential of big data [17]. This means that stored data is only helpful through analysis. Consequently, this initially rudimentary approach to a definition is outlined in a sub-complex way, but should not be deepened here as a basis for understanding. In summary, however, it can be postulated that according to the current state of research, the so-called ‘V’ describes the characteristics of big data. Now that Gartner’s 3 V model has been expanded to 5 V’s and onward to 7 V’s, we are reviewing ten characteristics of big data that are intended to define the term more precisely [18]: volume, velocity, variety, variability, veracity, validity, vulnerability, volatility, visualization, and value. Further analysis concerning the challenges related to term definition is found in the article “Big Data: An inventory and the ongoing search for a definition“. Further literature on the deeper origin of the term and its functionality and general potential has been sufficiently published.

1.2.2. Knowledge of the Logistics Structures

Concerning the research object to be examined, it should be emphasized that two independent subject areas come together. So far, logistics has been considered a sub-area of classic business administration with the other departments of economics as a cooperating unit, but it was essentially self-contained. The functioning of logistics thus provides information about which core processes are inherent and how these processes arise, run, and function. In this way, it is possible to identify potentials and areas of application and to assign them to specific processes in logistics [19]. To do this, it is necessary to understand the precise structure of logistics. A discussion of classic logistics, their origin, and knowledge of the structure is an essential cornerstone of the supplementary literature of this article. For example, Jünemann provides what is probably the most well-known definition of logistics [20]. The foundation of this area is the normative specifications of logistics. Based on this structure, additional operational elements such as the tasks, goals, instruments, and logistics potentials can be derived. This crystallization of a system creates the possibility of identifying processes. The knowledge gained in this way is always compared with relevant specialized literature, and differences are discussed. The topic is then systematized and explained from several perspectives.

The definition that Reinhardt Jünemann provided over 40 years ago shows other dimensions that make up the concept of logistics in addition to movement. The logistics of the 20th century increasingly formed into a complex subfield of business administration, which is becoming increasingly important in the context of digitalization. The cornerstones of this area are the normative requirements of logistics, which therefore means that:

• the right amount,
• the correct objects (including goods, people, energy, or information),
• in the right place in the system,
• at the right time,
• in the right quality,
• and at the correct cost to be provided [20].

This definition of logistics finds its framework in the various sub-areas, the interaction of which requires optimal management and control. The challenges posed to such subsystems’ management and control instruments have grown increasingly due to cost pressure and competition. For companies, efficiency and cost-effective processes are the answer to ever-shorter product life cycles that depend on the further acceleration of throughput times in production processes. These times face increasing coordination challenges due to the finely structured processes between the individual areas of logistics. For example, increasingly complex machines are being sold that require special training or instructions, regular maintenance, or, in the modern context, regular software updates [21]. At its core, for example, production logistics has a coordinative, flow-related function within the framework of production management, which must now also deal with sustainability aspects in addition to commercial production tasks [22]. In contrast, logistics structures and technical possibilities should be evaluated, as well as the current state of knowledge about new technologies (big data, digitization, etc.). Moreover, employees’ attitudes and value settings in logistics, process management, or similar job profiles should be assessed.

2. Literature Review

This article builds specifically on basic research in logistics and develops and/or reviews it based on existing literature. The logistics’ basic structure and the technology’s status quo are analyzed. The literature research was carried out systematically and followed the theses and goals of the article. A supplementary hermeneutically oriented approach is consequently assigned to the topic to be examined. Depending on the knowledge goal and the theses listed above, this leads to a research and expertise-related style and a reporting or statement-of-action approach (analytical).

Numerous scientific books and publications have appeared on logistics and big data and, whenever possible, have been used as primary sources. The existing literature is reviewed and analyzed. The hermeneutic approach is thus intended to illuminate the status quo of current scientific research in the field of logistics and therefore show the starting point for pointing out possible future developments.

The selected literature results from explicit knowledge of the topic to be worked on, a detailed overview of the existing literature, and other publications. The hermeneutic approach is limited to conference reports, journal articles, and selected specialist literature. Furthermore, the references of the identified papers are examined and critically assessed.

Monographs (e.g., Krieger or Jünemann) and collective works (e.g., Buxmann) were decisive for the article according to criteria of the subject area (here, basics of logistics, process management, and digitization) under the stipulation of topicality (the edition) and selected relevance. These standard works deal in detail with basic research in logistics and systematically illuminate the topic of digitization. For example, terms are delimited between individual definitions, such as digitization, automation, or big data [23]. Furthermore, these standard works are critically assessed regarding logistics structure and goals, as well as tasks and the effects of digitization at all levels.

In addition, publications from public institutions or registered associations (e.g., Bitkom e.V.) were used to compare the survey results. Finally, additional publications by the authors are used, which shed more light on the individual topics. Current articles from specialist journals and publications (in digital form, e.g., Kleemann) from companies were also used for big data. Finally, databases such as Google Scholar were used to supplement scientific documents.

3. Methodology, Hypotheses and Objectives

A quantitative empirical research method in the form of a survey was used in this article. The survey’s focus, which included 27 open and closed questions, was on evaluating the potentials, expectations, and dangers of dealing with big data technologies. The methodology for collecting information about the current status of the use of big data technologies followed the rules of empirical research along the quality criteria of reliability, validity, and objectivity. For the latter criterion, various industries were deliberately selected. In addition, the size of the company diversified, and logistics experts were explicitly surveyed. A total of 440 companies were included in the survey. A total of 150 companies answered the questionnaire, corresponding to a response rate of 34%. The questionnaire was evaluated using statistical methods.

When determining the sample size, a confidence level of 95%, a standard deviation of p = 0.5 (representing the maximum possible deviation), and an allowable amount of error of 8% were specified. For this purpose, the sample size must have a value of 150 companies. As mentioned, 150 companies answered the questionnaire, so the random sample can be considered sufficient. The formula and values to calculate the sample size are:

$$\mathrm{n}=\frac{{\mathrm{Z}}^2\ast \mathrm{p}\ast \left(1-\mathrm{p}\right)}{{\mathrm{e}}^2}$$

(1)

where,

• n = the size of the sample
• Z = Confidence value (here, 1.96)
• p = Standard deviation (here, 0.5)
• e = Quantity of errors (here, 8%)

It can be stated that a particular technical competence was required to answer the questions. For this reason, the respondents were selected in such a way that they had already come into contact with the subject of the study or, at minimum, had integrated big data applications into their everyday work.

First, it was determined to what extent decision-makers were familiar with big data in logistics. In addition, dependencies were also evaluated, such as the role of company size or industry.

The research was carried out using a deductive approach. A deductive method was preferred due to the diverse selection of literature in the relevant subject areas. Thus, the theses and objectives mentioned can be understood as an update of already known theories and can be individually adapted.

Qualitative research in the form of an expert interview was a supplementary research method to quantitative research. The interviews support the survey and substantiate the question. An excerpt of the results is presented in this article.

Furthermore, the content evaluation follows the model developed by Philipp Mayring for essential parts of the content analysis. The qualitative content analysis process took place in five steps, starting with selecting the materials, then determining the direction of the analysis and the form of the content analysis, interpreting the results, and ensuring the quality criteria [24].

The sample size calculation follows a population formula, the exact size of which is very large and cannot be precisely determined.

A graphical representation in the form of bars and pie charts supports the interpretation of the results.

Questions were asked based on a literature review and previous research (described in more detail in the introductory section). These questions were the basis for the formulation of the following research hypotheses:

Hypothesis 1.

The use of big data technologies in logistics exploits the optimization potential of processes and leads to an increase in efficiency in terms of logistics performance.

Hypothesis 2.

Logistics costs are reduced by using big data and artificial intelligence applications; investments pay for themselves in an economically reasonable period.

Hypothesis 3.

Due to logistics processes’ optimization and efficiency improvement, using big data entails an ecological paradigm shift consistent with climate goals.

Hypothesis 4.

The use of big data entails a social-ethical paradigm shift due to logistics processes’ optimization and efficiency improvement.

The empirical research aims to describe and present the most common instruments and methods of data science technologies in logistics.

Another goal of the survey is to provide information about the level of knowledge the experts have concerning big data technologies. This also includes the ecological and ethical perspectives associated with using such technologies.

Research activities usually consist of several steps. The first step was to examine the two topics of logistics and big data to identify the research problem and to find the research gap, which is reflected in the holistic approach. Unfortunately, there are few research papers devoted to this approach. Moreover, they deal with the topic only marginally or from a different research perspective. In the next step, the research areas (basic research for both subject areas) and the research goal were determined, leading to the establishment of the hypotheses mentioned above.

The data collected was based on a questionnaire about the topics.

The validation of questions was based on Cronbach’s alpha, since a repeated measure does not seem helpful.

4. Results and Discussion

4.1. General Questions about Big Data

To evaluate more precisely whether big data is not being misused as a synonym for digitization, the next step was to ask whether the terms ‘big data’ and ‘digitization’ are synonymous for the respondents. The term ‘big data’ is not clearly defined as such and is in the process of constant change as far as the descriptive attributes are concerned [16]. The term ‘digitization’ is often used synonymously with the term ‘automation’, which merely describes a work process processed by a task carrier (e.g., a machine) [25]. On the other hand, digitization only represents the translation of analog processes (including those of automation) into binary codes. For 14.81% of those surveyed, big data is simply a synonym for digitization. Interestingly, almost three quarters (72.73%) of subjects who answered “yes” to this question work in large companies. This indicates that the implementation of big data technologies is failing due to a financial hurdle and a lack of communication. The thorough examination of the topic still seems to have the potential for improvement in larger companies that are assumed to have a good financial background.

Questions that should classify the term big data are followed by a question that inquires about the extent to which the respondents are familiar with big data technologies in their professional environment. Figure 2 illustrates the results.

The study results allow a clear conclusion that familiarity with big data technologies changes depending on the company’s size. A total of 40% of the surveyed companies stated that they were sufficiently (17.78%) to fully (1.48%) familiar with the subject. There is a clear connection to the size of the company. If company size and the degree of expertise are presented in a context, the result is a correlation coefficient of r = −0.72. It can therefore be assumed that the larger the company is, the more familiar its employees are with handling big data technologies, or, as illustrated in the diagram here, the small proportion of those who can demonstrate excellent expert status in the field is reserved for larger companies. Therefore, there is a strong negative correlation. As a result, small and micro-enterprise expertise is underrepresented among the employees responsible for processes. Almost 44% of those surveyed who stated that they were thoroughly familiar with big data technologies in their professional environment come from companies with fewer than 50 employees or less than €10,000,000 in annual sales. Furthermore, the result of this question underpins the previous questions already discussed and shows once again that precise conceptual definitions and their communication appear necessary in the company.

4.2. Evaluation of the Potential of Big Data Technologies

Now that the status quo has been evaluated regarding the general terminology and the level of knowledge, the implementation in the company, and the understanding of the content of big data technologies, the results will now be discussed, which provide information about the potential. To this end, the employees were asked by experts on this topic various questions on identifying sources of error, assessments of decision-making processes, the potential for increasing efficiency in the context of energy and cost savings, logistical processes, and the necessary know-how.

4.2.1. Fault Identification

It should be noted that identifying sources of error in internal processes in the company is of significant importance for those responsible for the process. More than 60% agreed almost entirely with the thesis that big data technologies help identify sources of error in internal company processes. The aspect of error detection is not only a serious added value for reasons of efficiency, but also, from the point of view of sustainability, considerable potential can be tapped here. If sources of error are identified at an early stage, consequential damage that consumes resources can be avoided. Thus, preventing mistakes in process flows, random distribution, and unnecessary wear and tear has a lasting effect and is, therefore, energy and resource-saving. Furthermore, if this potential of big data technologies in connection with the Internet of Things (IoT) is consistently exploited, product life cycles can be extended again—an essential contribution to a more environmentally conscious and sustainable economy [26]. Figure 3 illustrates the results of this question.

From a sustainability point of view, the potential for detecting errors at an early stage cannot be underestimated. This primarily saves valuable resources and consequently avoids long transport routes. Raw materials, otherwise used for pure preservation of the substance of, for example, means of production, do not have to be laboriously mined and transported to the destination. Thus, the goal of sustainability and ecological compatibility and the classic goals, such as cost reduction, are served. All of this must be done with rebound effects in mind. Advanced technologies are undoubtedly capable of reducing emissions, for example. Overall, products can be manufactured more efficiently and thus with less use of resources. However, this also leads to an ever-faster demand for the latest product. A rebound effect is, therefore, a feedback mechanism that only partially or not at all realizes potential savings from increases in efficiency [27].

4.2.2. Decision Making

If big data technologies are essential and valuable for error detection for the aforementioned reasons, the interpretation of the following question should not necessarily be rated positively. Therefore, the following results as shown in Figure 4 have been evaluated on the question of whether big data technologies generally enable better solutions in decision-making processes than those created by people.

A closer look at the adjective ‘better’ is necessary when evaluating this question. When asked what they understand by ‘better’ decisions, the subjects’ answers were related almost entirely to economic choices. This question is based solely on an economic, efficiency-based, and profit-oriented value framework that largely ignores human, non-economic values. This raises the question of whether big data technologies also make better decisions that are more advantageous in terms of ethical humanitarian added value.

The question of efficiency is usually answered quickly. The subjects agree that big data technologies significantly accelerate decision-making processes. Especially in the field of collaborative work in logistics, digital solutions in connection with large amounts of data have promising potential. The possible applications have already been published in the article ‘Digital and collaborative solutions’ [28]. The approval took place with almost 70% (68.3%) of the respondents at large companies. The authors saw this potential more as an opportunity to improve communication channels resulting from the cross-evaluation of this question. The acceleration of decision-making processes is of secondary importance for smaller companies, since shorter communication channels are expected here due to the system.

4.2.3. Pattern Recognition

There is a consensus that big data technologies can recognize patterns. Pattern recognition is ultimately the core element that gives digitization its face in its application [29]. The necessary hardware as a prerequisite enables analog processes to be translated into zeros and ones. On the one hand, if digital technology (e.g., smartphones) is now used, every use leaves a digital footprint, which leaves behind vast amounts of data due to widespread use. On the other hand, patterns can be recognized in these amounts of data, and future behavior can be predicted from them. The subjects agree about this potential, with almost half fully agreeing. A concrete application area for big data technologies has been identified here, especially about recognizable patterns that indicate an ordering rhythm on the part of customers. In individual expert discussions after the survey, the people surveyed stated that this was a helpful tool for synchronizing the ordering rhythm in procurement with the ordering rhythm of the customers, i.e., synchronizing the supply with the demand. This makes logistical instruments, especially logistical software for pattern recognition, an indispensable management tool, especially in supply chain management. Thus, primarily the question of added value in the respective areas (here, in procurement) of logistics can be answered with the help of analytics through data science (here, pattern recognition) insofar as the ordering processes can be optimized concerning logistics performance. As a result, delivery times are shortened, logistics costs are reduced, and the holistic and customer-oriented perspective is sharpened. Explicitly used in the procurement process, big data technologies help meet classic logistics goals. At this point, however, no statement can be made as to whether what is described also contributes to achieving the new, expanded, already described target system.

4.2.4. Energy Saving Potential

The subject of the next question is whether energy can also be saved with the support of big data technologies through the expected increases in efficiency in internal company processes. Although most experts tend to agree with this statement (the average abscissa value is 5.9), a standard deviation of 1.65 indicates a certain degree of disagreement on this question (See below: Illustration). This statement is confirmed by answering the question that big data technologies lead to higher energy costs. More than half tended to agree with this statement, and around 10% (5.26% and 6.14%) agreed (scale level 7 to 8). These are the roughly 10% who also tend to believe that big data technologies have little potential for saving energy through increased efficiency. Consequently, it can be assumed that 10% of the surveyed experts are consistent in their statements and that a particular technical depth or knowledge of the potential and limits of big data can be assumed. The following Figure 5 illustrates the situation.

Conversely, this means that, in general, there needs to be more precise knowledge of the potential and dangers of big data technologies. In this context, particular attention should be paid to the threat emanating from the already-mentioned rebound effects. These effects are published in detail in the article ‘Ecological Digitalization’ [26]. At this point, training courses in handling big data technologies for process managers should be established as logistics standards. An ‘excess’ of production and consumer goods inevitably leads to an ‘excess’ of required resources and thus of energy, which probably cannot be compensated for by increases in efficiency. For example, large quantities of raw cobalt are needed to produce lithium-ion batteries, which are used in smartphones, among other things. Global demand is currently 110,000 tons per year and, according to calculations by the Dera raw materials agency, will be between 187,000 and 225,000 tons by 2026 [30]. These quantities not only have to be transported via logistical systems, but according to Jünemann’s definition, they also have to correspond to the guidelines of the logistical coordination and networking process [20].

4.3. Questionnaires on Ethical and Sustainable Aspects

To also evaluate the use of big data regarding the holistic system of objectives, the survey assessed whether there were any discrepancies regarding ethical and sustainability aspects. For example, to what extent are economical, efficiency-driven parts compatible with ethical and sustainable goals? First, two questions were asked, which aimed at the autonomy of process owners regarding decisions. In other words, do the decision-makers leave the decision-making process to AI?

Many process managers believe in this thesis or confirm this thesis. However, when asked whether big data technologies will essentially replace human decision-making in the future, there is disagreement about the total number of respondents (see Figure 6 left). The follow-up question that big data technologies undermine the integrity of employees or all target groups through the type of decision-making is also not straightforward. The slopes of the trend lines of both graphs even tend to be slightly negative.

The interpretation of the answers to this question and the help of the individual expert discussions allow the conclusion that the decision-makers consider it possible from a technological point of view to leave the decision-making process to algorithms, but this is not the desirable case. The reason for this could be a complex value system of human psychology, which cannot be found so extensively in repetitive patterns. This clearly shows that a decision-making process is not only made based on rational economic concerns, but that social and ecological values also flow into the decision-making process. In conclusion, decision-making using big data technologies upholds the integrity of employees and all target groups. It remains more of a wish than the father of the idea. For example, implementing big data technologies is likely to be hampered by a certain lack of acceptance by employees.

The increasing use of cobots is an example of incomplete holistic implementation involving ethical and sustainable issues. Especially in joint production or service companies, it is essential to analyze the employees’ work environment requirements and to align them correctly in terms of value. This orientation considers employees who encounter communal facilities, so-called cobots, and share their workplace with them. A cobot is generally understood to be a collaborating lightweight robot that can be operated intuitively by the employee and can be integrated into existing production. To do this, certain conditions must be met. A cobot must be adapted to the respective employee, i.e., set up. As a result, user-adapted programming takes place, but is initially aimed at the work to be performed. When implementing cobots, the question of employee acceptance should be addressed.

The solution could be implementing integrated management systems that focus on increasing the acceptance of cobots in a shared workplace, i.e., letting the cobot and the human become ‘friends’. There is a need to find a sustainable way of working together to make the common space safe and environmentally friendly. Therefore, implementing integrated management systems should be carried out according to ISO standards (ISO 45001 and 14001) [31]. Furthermore, implementing the processes within the framework of ISO 9001 and ISO 10007 is recommended [32]. Finally, working out a win-win strategy is recommended within the ISO 10014, 10015, and 10018 [33].

So, suppose doubts remain about the autonomy of those responsible for the process. In that case, there is, again, great agreement as to what the security and trustworthiness of the data means to customers, suppliers, employees, and target groups. The average was 2.96 with a standard deviation of 1.21 of the abscissa value. Less than a third of those questioned find that this trustworthiness is only sufficient or even insufficient. None of the surveyed experts gave the school grade “6”, i.e., insufficient. When cross-evaluating this question, it is noteworthy that the skepticism comes mainly from large companies (62.5%). This contradicts the expectation that, for budget reasons, there should be better ways of meeting security standards and remedying security deficiencies, especially in large companies. This is also confirmed by a study for IT security in Germany after this market has been growing continuously and reached a new all-time high of 6.2 billion euros in 2021 [34]. The sales figures and market forecasts for the IT security market in Germany are based on current calculations and studies by the IT market research company IDC. In addition to public institutions, the demand for IT security tools was mainly generated by large companies.

4.4. Qualitative and Quantitative Requirements

Concerning the quantitative requirements, whether relevant data is available in the company should be examined. When asked what percentage of all the necessary data/information was currently available, almost three quarters (73.22%) of all surveyed experts answered that more than half of the relevant data was available. At 44.65%, more than 75% of all relevant data is available. According to the subjects, relevant data means all the data usually used for business analysis (see Figure 7).

The trend towards increasing networking is being strengthened and has kept momentum so far. The values from the Bitkom survey from 2018 were far exceeded four years later [35]. It can therefore be concluded that digitization levels 0, 1, and 2 have been fully achieved. The proportion of digitized companies in the industry is significantly lower than, for example, in the service sector. The potential in these areas is correspondingly high. This also means that companies in the industrial sector have caught up with service companies. The translation of the analogue world into binary codes is complete, at least for the German-speaking area. Large companies also tend to have an advantage regarding data availability. Eighty percent of subjects that indicated that more than three quarters of the relevant data were available were significant (64%) and medium-sized companies (16%). Only 18% of small companies and 2% of micro-enterprises have more than 75% of the relevant data available. The industry distribution is also attractive, as it provides information about the potential of using big data technologies. The responses about data availability from the automotive, IT, and construction industries were consistently positive. The automotive industry (14%), followed by the IT sector and the construction industry (12% each), it was most frequently agreed that more than 75% of all relevant data for business decisions were available (see Figure 8).

The cross-evaluation shows that, conversely, there are also sectors that lag in data availability compared to the automotive industry, for example. Using big data technologies would be desirable in the pharmaceutical industry since social problems could also be solved quickly here. The production of medicines or vaccines is directly related to the metadata of human existence or depends on it. For example, areas of application of data science technologies could help to answer the question of when and how a pandemic spreads. This helps develop a suitable drug and significantly speeds up the process. Industries in the social sector, as well as the hotel industry, the real estate industry, or the water, sewage, and waste disposal industry are represented in an under-complex manner regarding data availability. Logistics in the public sector in particular could benefit from big data technologies here, which is reflected in the reduction in complexity. As a result, this leads to an increase in the operational excellence of customer-oriented and internal processes at the micro level in politics or regional associations.

For the use of big data technologies, industry-specific investments in the progressive training of employees are therefore necessary. In 2020, Bitkom stated in a study they conducted that 70% of the 1104 companies surveyed (20 or more employees) send their employees to further training to prepare them accordingly. The information on investment in further training varied based on the number of employees. While the proportion of subjects with more than 499 employees who invest in additional training was as high as 78 percent, the corresponding ratio of companies with fewer than 100 employees was around 9 percentage points less [36]. A certain level of competence in dealing with data and digitization, in general, is essential to companies. However, the hoped-for added value still needs to be determined. Due to the complexity that big data and, above all, the data analysis entails, future fields of application can be estimated, but the effects can hardly be measured [37].

There is disagreement as to whether there is a need to acquire special knowledge in dealing with big data technologies. A total of 50% of subjects consider comprehensive specialist knowledge to be indispensable, and 43% counter that general IT knowledge is sufficient. Good IT skills mean safely handling specific application programs such as Microsoft Office or Photoshop. This question did not consider IT specialists who can develop professional programming skills. A total of 7% of the experts could not assess this question.

Currently, 12.5% of the data is collected using exclusively analogue methods (e.g., registration via files, lists, etc.). Furthermore, 65.18% of those surveyed use the usual MS Office programs or similar software from other providers such as Apple. In addition to the necessary financial means, implementing new technologies in the company also requires the willingness to accept changes. Change management offers a sufficiently large range of instruments for this and also describes the various phases of a change process [38].

In addition to answering the question regarding the technologies used so far, the experts were asked how data is generally collected. According to the cross-evaluation, around a third (32.14%) of subjects stated that the employees recorded the data almost exclusively manually (see Figure 9). Without exception, all respondents who chose this answer option stated at the same time that they were familiar with the term ‘big data’ and that half of this group was also well acquainted with big data technologies.

This essentially allows for two theses:

• There is disagreement about an exact and uniform definition of big data in general and its applications in particular.
• The implementation of big data technologies is unprofitable for economic reasons.

The first thesis can be justified by sometimes contradictory statements regarding the definition, areas of application, or potential, which can only be explained by the fact that the theoretical basics are well-known but are not applied in the practical course of the process. Nevertheless, almost half of all respondents agree that patterns can be recognized, and future behavior can be predicted.

The second hypothesis needs to be examined in more detail. However, if knowledge of the potential and possible applications is assumed to a certain extent, it is reasonable to assume that the implementation of big data technologies in companies will fail due to economic aspects.

The interpretation of the evaluation of this question requires a closer look at the economic level. Thus, at least the aspect of economic efficiency should be examined more closely to better understand the use or the renunciation of these technologies.

4.5. Technological Requirements of Big Data Solutions

One of the core requirements for applying big data technologies is defined by the fact that an object can be located. This does not only mean the means of transport, but the primary good as such. Therefore, the next question for the experts is whether it is even possible to locate objects in supply chain processes, both in external logistics and intralogistics. Figure 10 shows the results.

Almost a quarter (22.52%) of subjects stated that they were unfamiliar with the location of objects, i.e., they did not know whether the technological possibilities would allow objects to be located separately. According to the cross-evaluation, this ‘knowledge gap’ can be traced back to almost 70% (69.23%) large and nearly 20% (19.23%) medium-sized companies. A brief summary of this value is that workshops with a special focus on locating projects and the associated measured variables in the ongoing process should be given to those responsible for the project and process. Every responsible employee should be aware of the existence of an essential building block for the application of big data technologies.

For a little more than half (cumulatively 56.75%) of all companies, the state of the art is that at least A-articles can be located in connection with the transport system. This also corresponds to the expectations regarding the digitization level, in which companies are classified according to their digitization progress [28]. Companies that achieved at least digitization level 2 can locate at a minimum A-items in logistical processes. This localization ability is a mandatory prerequisite for optimizing the flow of materials and, therefore, for optimizing the logistic performance along the logistic target system.

However, other target logistic values can be optimized in the sense of localization. In this way, unnecessary transport routes can be avoided. Because it is possible to see an object’s location at any time, reactions to emergencies or other unexpected events (vehicle breakdown, accident, traffic jam, or detour) are much faster. In addition to saving time, resources are also reduced. As a result, achieving and implementing sustainability goals in logistics is easier. Both economic and ecological objectives can be pursued in C-parts management. In an industrial company, C-parts are required in various forms. They, therefore, form a diversified and small-scale portfolio of complex individual parts with even more complex procurement channels. C-parts have the characteristic that the procurement effort is disproportionately high compared to the contribution to value creation. The purchasing volume usually corresponds to only a small part of the total volume. Sourcing, organizing repeat orders, and quality control can be defined as complex processes in the company. To save costs and resources, it is necessary to record C-parts systematically. With the localization ability of objects as a prerequisite and the appropriate software (AI), new procurement routes can be determined, the wealth of product data can be structured, and tailor-made transport routes can be found for a smooth flow of materials. Almost half of the subjects do not know where C objects, in particular, are in the process, which means that there is great potential for object localization optimization. Training courses on the technological possibilities, the costs of these technologies, and their application areas are urgently recommended.

4.6. Economic Aspects of Implementing Big Data Solutions

For the economic assessment of the implementation of big data technologies, the experts were particularly asked about the profitability of big data solutions. The survey question was whether the investments in big data solutions would pay for themselves after three years (including personnel costs). This question requires a basic understanding of business know-how, which can be assumed due to the respondents’ respective professional positions and the associated requirements. Economically speaking, more than half (56.37%) of respondents tend to agree that they will pay for themselves after three years (cumulative value of the dark blue bars in Figure 11). In most cases, the expert interviews were based on a dynamic investment calculation, which discounts the future payments from the initial investment to the respective cash values. However, the correctness and plausibility of these calculations should have been checked during the expert interviews.

The experts’ statements showed that the capital values of investments in big data technologies were primarily positive in large companies. Economically, in the sense of the profitability of an investment in big data technologies, it can be stated that the implementation is or will be worth it in monetary terms if the output is high enough, which is associated with the effect of a fixed cost degression. These economies of scale are often only reserved for large industrial companies, such as the automotive industry. The cross-analysis confirms this and shows that more than three quarters (77.78%) of the dark blue bars are attributable to large companies.

With regard to the use of big data technologies, the experts were also asked what would speak against an introduction in the company.

The evaluation showed that 43.93% of all respondents stated that a lack of technical IT competence was the reason for the resistance to implementing big data technologies. The willingness to change in the company also plays a major role. According to the study, the lack of acceptance of innovations is one of the main reasons big data solutions have encountered internal resistance so far. As expected, financial resistance is also an obstacle to introducing the technology at all stages of expansion.

The statements regarding the economic arguments are also reliable because big data technologies are already being used in almost half of all surveyed companies. Initially, only use time was evaluated. The extent of use or the company’s area where these technologies are used were not evaluated. Nevertheless, 22.73% of companies stated that they had been using big data technologies for at least a year, and 20.35% said they had used it for at least three years. The following Figure 12 explains the situation.

Among the companies already using big data technologies, almost 80% are large companies (purple bars). This confirms the statement made above. At this point, it makes sense to differentiate according to the company’s size, assign the level of knowledge and potential, and thus develop targeted further training measures. These training courses should clarify economic and technical questions and include assessing the consequences of using big data technologies. This necessity is derived from the extended target logistics system (as already mentioned). This also seems necessary, as 41.12% of those surveyed plan to implement big data technologies in the company in the next three years. Ethical questions and sustainability aspects, which are inextricably linked to these technologies, are becoming urgent and require a concept so that the logistical target system can be penetrated holistically.

5. Conclusions and Recommendations for Action

If knowledge of the potential and the technical knowledge required for the use of large companies is assumed and partially proven, the cause-and-effect principle remains unanswered. A frequently occurring problem raises the question of which criteria should be used to identify patterns from which relationships (correlations) can then provide information about certain facts using statistical methods. Artificial neural networks are usually understood as a ‘black box’. The traceability of the criteria according to patterns is difficult or impossible, even for the programmers of these algorithms. In this way, new patterns are formed from created patterns, which means that the underlying cause of the decision can no longer be identified—the ‘why’ of the decision remains unanswered and consequently leads to ethical questions [39].

The fact that the cause-and-effect technology is largely understood as a black box even by professional IT experts shows that the assessment of subjects requires a more detailed information analysis on training and further education in the field of big data. Technical aspects should be taken into account as well as an ethical interpretation of the results from big data analyses. Special knowledge should therefore be conveyed in a multidimensional manner. The evaluation of the survey in general in connection with this specific question allows the following conclusions with a classification of the expandability (based on a school grading system, which stands for 1 = fully represented to 5 = strongly in need) (see

Table 1. Level of knowledge assessed according to school grades (1 = very good, 5 = insufficient).

Area of Knowledge	Grade
Technical, user-specific skills in dealing with big data technologies (industry-independent)	3.5
Knowledge of the areas of application, potential, and possible benefits of big data technologies in general	2.0
Knowledge of the potential and possible benefits of big data technologies in the logistics sector	4.0
Knowledge of the economic potential and possible benefits of big data technologies	3.0
Knowledge of the security of data and possible legal distortions (data misuse)	1.5
Knowledge of the cause-and-effect principle of big data technologies in the context of ethical issues
Knowledge of energy consumption and the environmental impact of using big data technologies	3.0

). This assessment is largely based on the cross-evaluation of the questions and their partially deviating/contradictory answers. Due to the limited data situation and incomplete information, the values shown in the second column are to be understood as heuristics.

The lack of knowledge about concrete potentials, especially in the logistical and ethical area, also allows for the conclusion that a differentiated view of business models in connection with big data technologies is represented under-complex. The survey data confirms this conclusion. A little more than a third of the experts see concrete internal employment opportunities in logistics, and especially in procurement (see Figure 13).

The possibilities in warehouse logistics and transport, in particular, seem to be expandable in terms of efficiency once the necessary technical prerequisites have been created. Enhanced Process Automation (the area of application includes processes for which no standardized and structured data must be available) could ensure more resource-efficient handling in transport logistics, but also warehouse logistics, by shortening transport routes, or better yet, superfluous routes are identified and eliminated. In this way, intelligent tour planning can be used to act more sustainably and improve a company’s or public institution’s profitability.

In procurement logistics, the added value could be achieved by optimizing procurement processes by focusing more on C-parts management. Goods with a low individual value but a high procurement quantity are generally defined as C-parts. The palette ranges from office supplies to tools, which is less important from a monetary and strategic point of view. Therefore, there is also unused savings potential outside of procurement logistics that could be exploited with the help of Big Data technologies. The costs incurred in procurement processes can be significantly reduced through the targeted use in C-parts management.

In the individual expert interviews, those responsible for the process stated that the proportion of the total costs for C-parts is represented by approximately 80% of process costs. Accordingly, only about 20% can be assigned to the material. The high proportion of process costs can be explained by a wide variety of articles in addition to the small order sizes and the low material unit price. This means that the individual work steps in procurement, from determining requirements to document entry, generate unstructured, individualized and, above all, enormous data streams. These work steps, which proportionately have a much more significant influence on the overall costs than the pure material costs, can now be optimized with the aid of the technologies mentioned [40].

Given the challenges the 21st century has in store, the survey result for this question sets new tasks for clarifying the potential of big data technologies in companies. However, unfortunately, virtually no potential is recognized in disposal logistics. Based on the results of the previous questions, summarized in the table above, it is hardly surprising that the added value of using artificial intelligence, especially in waste disposal logistics, is not discussed in a complex way. However, this discourse is essential to counteract the environmental problem, at least when it comes to disposal, with good concepts.

A concrete application of AI in disposal logistics can currently be found at the municipal level. The city of Speyer has been using it since 6th May 2021 with a camera controlled by AI-based software on the city cleaning sweepers. The pilot project ‘Digitization in the field of city cleaning’ should enable an exact classification of waste and thus detect so-called waste hotspots [41]. This concept could also become more critical in a business context since various waste products are also produced during the production process, which must (or should) be separated and disposed of afterwards in a time-consuming and cost-intensive manner. Furthermore, artificial intelligence can analyze valuable ingredients during disposal, thus contributing to the desired circular economy. Resource consumption and emissions, which result from the manufacture, use, and disposal of hardware, can and must be counted among the direct environmental impacts in this context. If the new technology is applied, the induced changes in production patterns can have indirect ecological effects, which are reflected in the disposal costs [3].

These costs are not included in the production costs; they are externalized. This means that there needs to be more implementation in the prices of the products, which would be considerably more expensive. The internalization of external costs is, therefore, an important step to create incentives within the business logic framework to (re)localize production and thus avoid long transport routes. Concepts of sustainability, for example, through product developments that extend the service life and could therefore justify a higher price for the customer, are a starting point in economics [26]. Unfortunately, the acceleration and disruptive character caused by digitization often appear to be contrary to the values of our culture. However, this can be separate from efficiency thinking and profit maximization. A survey of several hundred executives in Germany conducted by the “Values Commission–Initiative Values-Conscious Leadership e.V.” shows that value-oriented leadership is “very high” for a credible value system for corporate success. Over 90% of managers agreed with this statement [42]. It is important to implement these values in the logistical target system. In addition, a clear work assignment in the ethical and economic sense for computer scientists can be recorded. The computer scientist is responsible for all the consequences that his actions trigger directly or indirectly. i.e., For what others do or can do with the products he has developed or concepts he has devised [43].

With the help of ethical values, the future of modern, digitized logistics can once again serve the purpose of society and, if applied correctly, mean real progress for the community. This requires concrete action in the company because of the implementation of the ethical visions and economic objectives that have already been conceived. In this way, it is possible to reconcile business and social values and establish them sustainably.

One of the goals of the research questions was to define a meaningful subdivision of the contact groups. The identified potentials and problems should then be assigned to the relevant stakeholders. The survey results identify four potential interest groups that could benefit from the findings and take targeted action accordingly. The four dimensions are shown in Figure 14 with an example.

The results essentially allow a subdivision into four problem areas, which require a need for action in the direction of improvement/optimization in the respective dimension (see Figure 14):

• Knowledge of economic and application possibilities of big data technologies,
• Knowledge of ecological effects when using big data technologies,
• Ethical problems in dealing with big data technologies,
• Technical knowledge in dealing with big data technologies.

For example, some concrete concepts have been published by the Federal Ministry for Economic Cooperation and Development and must be assigned to the respective dimensions [44]. The description of the dimensions is as follows:

1st Dimension: Within the political framework, legislation on accounting for external costs could be considered at the national level, for example. These costs also include disposal costs. A law that puts a sharper focus on disposal logistics, i.e., clearly defining the transport routes and responsibilities for disposal, could encourage companies to develop better disposal concepts. Regarding this aspect, the companies saw little potential for using big data technologies (see Figure 13). The topic of externalization was taken up in the article ‘Ecological Digitalization-Potentials, Limits, Dangers’ [26]. Furthermore, the decision-makers in companies need to be sufficiently aware of the effects of big data technologies and their possible applications. Politicians could help with targeted information campaigns to bring transparency to the disposal processes. In addition, monetary incentives (e.g., through tax benefits) could encourage companies to organize their disposal logistics more efficiently.

2nd Dimension: To better meet the goal of sustainability, companies could pursue collaborative approaches in the future. The partly insufficient knowledge about the technical possibilities of Big Data technologies, especially in networking, means logically untapped potential with regard to the joint use of resources and, as a result, resource conservation. Concrete recommendations for action are published in the article ‘Digital and collaborative solutions’ [28].

3rd Dimension: This level is aimed at globalized cooperation, considering ecological and ethical framework conditions. This requires a conscious selection of suppliers and transparent supply chains. Big Data technologies could support the creation of concepts that harmonize the various laws, standards, and guidelines of the countries involved.

4th Dimension: Companies should consider their decisions in relation to all target parameters of logistics, especially if these were made by the algorithms they use. The decision-maker should be able to explain the main criteria according to which an algorithm calculates an output. This is a challenge since the decision-making structure in deep learning applications can no longer be understood even by software developers due to the complex software architecture of artificial intelligence. At this point, it is necessary to examine ethically and legally whether using such an AI is even permissible in a certain area. For example, selecting a supplier with the help of AI-supported software and under an exclusively economic value ethic could lead to wrong decisions (see Figure 4).

When an algorithmic decision-making process has been completed and accepted, decision-makers should be able to explain this in terms of the algorithm or human decision (which may differ from the algorithm). Here, sustainability is important and should be communicated to politics and society to generate a continuous improvement process. This is an essential step towards transparency for the public.

The research questions asked at the beginning can be answered using the survey results and the secondary research. In particular, the first question about the potential and applications could be fully answered in production and warehouse logistics. At the same time, a possibility in the field of disposal logistics should be ‘awakened’. Necessary measures are only developed and implemented within the company if decision-makers bring a certain degree of sensitivity to a topic. The survey clearly showed that this sensitivity is under-represented.

There is usually disagreement among the responsible experts when identifying the problem areas. However, there is consensus when it comes to data protection and data security. Identifying dangers and problem areas on an ecological and ethical level is necessary for mastering these future challenges. For this purpose, the recommendation is to develop coordinated concepts in all four dimensions, which contain, communicate, and implement the topic ‘ethics and sustainability of Big Data technologies’. Workshops, events, trade fairs, campaigns, or university lectures could be practical tools in all dimensions.

Regarding the added value of big data technologies, it can be stated that from an economic point of view, the questions about capital values and returns from and to these technologies can be answered precisely, but are currently mostly reserved for large companies. Entrepreneurial concepts will also find their way into small and medium-sized companies in the future. Determining the added value from an ecological and ethical point of view requires further research. This research should begin with identifying the parameters that measure this value; in other words, posing the question of values that goes far beyond the economic point of view. Here it is recommended (also in every dimension) to set up a permanent ethics committee that answers questions about sustainability. The commissions could communicate, use common resources, and create concepts collaboratively.

Formulating concrete concepts, regulations, or even laws makes further research in this area necessary. Further research is also necessary to adequately answer the research questions and to close the research gap described.