Potential analysis of track-and-trace systems in the outbound logistics of a Swedish retailer

Potential analysis of track-and-trace

systems in the outbound logistics of

a Swedish retailer

MASTER-DEGREE THESIS WITHIN: Major NUMBER OF CREDITS:

PROGRAMME OF STUDY: International Logistics & Supply Chain Management

AUTHOR: Daniel Goll & Nils-Ole Bolte JÖNKÖPING May 2020

i

Master Thesis in Business Administration

Title: Potential analysis of track-and-trace systems in the outbound logistics of a Swedish retailer

Authors: Daniel Christopher Goll and Nils-Ole Bolte Tutor: Susanne Hertz

Date: 2020-05-18

Key terms: Outbound logistics, Track-and-trace technologies, Swedish retailer, Process optimization

Abstract

Supply chain visibility has become a crucial factor for companies in times of globalization and customer satisfaction. Track-and-trace technologies are important tools in order to enhance supply chain visibility. This thesis was written in cooperation with a Swedish retailer and evaluates potential track-and-trace technologies in order to develop a solution to close their current track-and-trace gap in their outbound logistics. Currently the handover point between the retailer and the postal service provider is not clearly defined, so that shipments get lost during the transition. Therefore, a literature review about currently used track-and-trace technologies was carried out. Several technologies with a wide price and applicability range are currently used and have been analysed regarding their strength and weaknesses. A qualitative study in form of interviews was conducted within the Swedish market about how this gap could possibly be closed. Empirical findings show that the existing track-and-trace technologies do not provide a best practice solution. Especially in the field of outbound logistics, several factors and the individual process requirements of a company have to be considered in order to develop an efficient solution, so that the existing track-and-trace gap can be closed. Each company has its unique set of challenges, which have to be solved in order to successfully implement a long-lasting tracking solution. A high dependence from the postal service provider is additionally given since all process steps need to be aligned to guarantee reliability of the data afterwards. In the case of the Swedish retailer, an automatized scanning bow with a separated area for outbound parcels is expected to improve transparency of the handover and lower the total amount of lost shipments. The breakeven point would be reached within the next years, so that operational saving could soon be achieved. Due to the global outbreak of COVID-19, as well as significant problems of the retailer, the practical application could not be tested. It should therefore be part of further academic studies.

ii Table of contents

1.

Introduction ... 1

1.2 Background ... 1

1.2 Problem discussion ... 2

1.3 Purpose and research question ... 4

1.4 Scope & Delimitations ... 5

1.5 Outline ... 6

2.

Literature Review ... 8

2.1 Definition of track-and-trace in logistics ... 8

2.2 Technical foundations of track-and-trace systems in logistics ... 9

2.3 Basics of position determination ... 10

2.4 Auto-ID systems in logistics ... 11

2.5 Outdoor positioning technologies in logistics ... 12

2.5.1 GPS ... 12

2.5.2 GSM ... 14

2.6 Indoor positioning technologies in logistics ... 16

2.6.1 Barcode ... 16

2.6.2 RFID ... 19

2.6.3 W-LAN ... 20

2.6.4 Ultra-Wide Band ... 22

2.6.5 Bluetooth Low Energy ... 23

2.7 Implementation factors of outbound tracking technologies ... 24

2.8 Empirical model of business process modelling ... 26

2.9 Basic architecture of tracking networks ... 27

2.10 Summary ... 28

3.

Research methodology ... 30

3.7 Case selection & sampling ... 34

3.8 Data collection ... 36

3.8.1 Secondary data collection ... 36

3.8.2 Primary data collection ... 36

3.9 Impact of COVID-19 ... 37

3.10 Data analysis ... 37

3.11 Research quality ... 38

4.

Empirical findings ... 40

4.1 Empirical model of business process modelling ... 40

iii

4.3 Current track-and-trace processes of Intersport ... 41

4.4 Outbound processes of Intersport ... 42

4.5 PostNord – The linking logistical service provider ... 43

4.6 Outbound processes of various retailer ... 43

4.7 Summary of outbound logistical processes ... 46

5.

Analysis ... 49

5.1 Evaluation of the examined track-and-trace technologies ... 49

5.2 Potential solutions for Intersport ... 53

5.3 Implementation recommendation for Intersport ... 58

5.4 Proposed solution for analyzed cases ... 61

6.

Conclusion ... 63

7.

Discussion ... 66

7.1 Implications ... 66

7.1.1 Theoretical implications ... 66

7.1.2 Practical implications ... 66

7.2 Limitations and future research gaps ... 67

iv Figures

Figure 1: Basic Architecture of Tracking Network ...27

Figure 2: BPM of current outbound logistics process ...42

Figure 3: Modelled solution with scanning bow ...53

Figure 4: BPM of the outbound logistics with a scanning bow ...57

Tables Table 1: Summary of track-and-trace technologies ...29

Table 2: Information about interview partners ...35

Table 3: Summary of analysed cases ...48

Table 4: Cost calculation of proposed solutions ...55

Appendix Appendix 1: GDPR Thesis Study Consent Form ... 77

Appendix 2: Coding process example ... 80

Appendix 3: Interview questionnaire with Swedish retailer ... 81

1 1. Introduction

______________________________________________________________________

The first chapter presents the background of the thesis, providing a general understanding of the topic of tracking and tracing the supply chain context. Following, the underlying problem is brought up leading up to the purpose and the research questions of the thesis. Lastly, the scope and delimitations are discussed.

______________________________________________________________________ 1.2 Background

In times of globalization, companies are increasingly required to become more flexible and agile in order to be able to meet the increasing adaptation needs and fluctuations in demand of customers (Klein & Thomas, 2009). Moreover, outsourcing has become another trend which requires the involvement of more entities in the supply chain. Lastly, product life cycles become shorter, which results in a lack of historical data and reducing organizations forecasting ability. Therefore, the topic of supply chain transparency for companies has become increasingly important in recent years. Supply chain transparency is the ability to track a wide variety of goods during transport in order to obtain a clear overview of the inventory and activities of the individual process steps. It enables companies to improve their customer service and cost control by managing inventory in motion, proactively updating status, limiting disruptions and mitigating risk (Gnimpieba, Nait-Sidi-Moh., Durand & Fortin, 2015). This allows companies to make their supply chain more agile and transparent to meet the demands that arise (Yao, Wu, Wang & Soc, 2009). Companies increasingly face problems with efficiently tracking and tracing their products within their logistics supply networks. This can lead to coordination problems throughout the supply chain and customer dissatisfaction. Increasing supply chain transparency through appropriate systems enables the identification of the location of the shipment so that the customer is able to trace the delivery process. Without suitable systems, companies face enormous challenges as it is almost impossible to ensure the required supply chain transparency (Baresi, Meroni & Plebani, 2016). Keeping track of all aspects of the supply chain is a challenge, especially as it can be difficult to control. The lack of anticipatory demand visibility leads to increased inventory levels at various points in the supply chain (Sun, 2017). In addition, lack of supply chain visibility can prevent companies from filtering relevant data in their supply chain. Thus, they cannot

2

make intelligent business decisions, anticipate potential challenges, and respond in a less random manner, as well as create unity throughout the supply chain and improve collaboration to deliver products in time (Baresi et al., 2016). Increased supply chain visibility, however, can lead to an improved workflow and increased profits. It enables companies to lower costs for customers and reduce risks (Do, Anke & Hackenbroich 2006). One of the key benefits is increased agility and flexibility. Increased visibility allows companies to focus on managing a supply chain and to make the appropriate changes to increase productivity regardless of circumstances. Increased agility also results in increased speed. Real-time, data-driven decisions enable business decisions to be made faster and risks to be minimized. This also enables the company to optimally meet increasing customer needs and to shift and adapt supply chains accordingly (Shamsuzzoha, Ehrs, Addo-Tenkorang, Nguyen & Helo, 2013). There are several ways to increase supply chain transparency. Research has shown that tracking and tracing has become one of the most important key technologies to meet customer needs and manage logistics networks efficiently (Lin & Zheng, 2013). According to Jagwani & Kumar, (2018), the most important technology investments, that should be made in the logistics industry over the next five years are Big Data and predictive analytics. Track-and-trace technologies are important enablers to generate big data and to effectively predict and analyse data in the supply chain. Without an adequate and suitable tracking system, efficient coordination of the logistical processes within a supply chain would be difficult to achieve. By implementing this system, it is possible to identify all uneven situations in the supply chain and make intelligent business decisions based on real-time data (Klein et al., 2009). In addition, tracking and tracing systems can be used to meet increasing customer demands regarding delivery status and position tracking of the respective shipments (Jagwani & Kumar, 2018). Therefore, it can be concluded that tracking and tracing systems offer a wide range of capabilities and functions to meet the growing demand for supply chain visibility (Klein et al., 2009).

1.2 Problem discussion

As described shortly in the introduction, the core of the problem is based on the handover points of the goods in the supply chain. Due to the lack of communication abilities between different systems, goods can get lost at points of transition once they are already passed on by one system but not registered within the system of the subsequent handling

3

partner. This leads to a permanent break regarding the visibility of each product along the supply chain. In addition, the cost side must be considered, because tracking gaps and parcel losses lead to unnecessary costs such as re-scanning, value replacement and handling costs. Sometimes companies try to overcome these issues by implementing additional systems, which should fill the gap and bring clarity at the handover points. When analysing the supply chain, it becomes evident that this might be beneficial for a single entity, but looking at the coordination between all entities, these steps are inefficiencies and should be limited (Gnimpieba et al., 2015). Specifically, the field of outbound logistics is prone to the mentioned coordination problems because it is the interface between the internal and external environment. Usually an external logistics service provider takes over the shipments and a clear handover point in terms of data transfer between the entities is necessary. Unclear data transfer and handover points can lead to track-and-trace gaps and lack of the overall supply chain visibility. Thus, one of the central challenges is the lack of the possibility to exchange information and data. If that is not given, the problem arises that companies cannot visualize their global systems in real time and provide the required information at any time. In addition, companies can lose control over the inbound delivery network due to the lack of visibility (Sun, 2017). Therefore, the challenge is to create a tracking system that is beneficial for both parties. It is important that both sides accept the system as the reference in order to prevent conflicts about the meaningfulness of the collected data. If one side puts the accuracy of the system or the collecting procedure in doubt, the data stored is no longer applicable for argumentative purposes between two parties. A clear proof of handover cannot be longer achieved (Zhou & Benton, 2007).

To implement such a system, trust between the involved parties must be developed. The sharing of data is a crucial step as this allows the handling partners within the supply chain a more detailed view into processes. However, it ultimately leads to an improved overall performance of the supply chain. It also manifests the cooperation between the involved companies, since substituting an involved company requires time and effort in order to replace them with a relatable company. It also contains the investment of financial resources of the parties to develop such an aligned system, which further emphasizes the commitment of the companies to the individual supply chain and its future well-being (Kocoglu, Imamoglu, Ince & Keskin, 2011). Consequently, the problem is a

4

multicomplex issue with several factors that must be regarded when implementing a track-and-trace technology. If implemented correctly, it decreases duplication of work, handling costs and increases transparency and safety for the goods at the point of transition. It also enables a better ad-hoc planning once sudden problems occur. All these benefits require the willingness of all related parties to cooperate. These issues lead to the focus of this study, which is put on a specific retailing company in the Swedish market regarding their outbound logistics processes. They are affected by the previously mentioned problems and require an in-depth research in order to develop a custom-made solution.

1.3 Purpose and research question

The individuality of the requirements within a company as well as in the respective supply chain shows that track-and-trace systems cannot be implemented in a generalized way. The literature to date has focused largely on possible track-and-trace systems specifically for indoor and outdoor localisation. The field of outbound logistics is rarely considered, specifically in this context. Given the gap identified in the literature, research with specific emphasis on track-and-trace technologies in the outbound logistics is rarely conducted. Furthermore, this study aims to solve current track-and-trace challenges of a Swedish retailer. Hence, the general purpose of this thesis is as follows:

“To evaluate the most important track-and-trace technologies in the field of outbound logistics and how they can effectively be implemented into a Swedish retailing company, in order to overcome current challenges.”

To reach the purpose, three research questions were created to close the existing research gap. The first part of fulfilling the purpose is to examine the current trends and technologies of track-and-trace in the field of outbound logistics. Thus, the first research question of this thesis is:

RQ1: What are current trends and techniques for track-and-trace in the supply chain context of outbound logistics?

5

The first research question is supposed to be answered within an extensive study of the existing literature. Given the required informational background about track-and-trace technologies, it is important to examine which factors and parameters have to be considered when implementing a track-and-trace solution. Thus, the second research question of this thesis is:

RQ2: Which factors have to be considered when implementing track-and-trace technologies in the outbound logistics of Swedish retailer?

Given this informational background it has to be evaluated how these technologies can be implemented into the individual context of the outbound logistics of a Swedish retailing company. Thus, the third research question of this thesis is:

RQ3: How can track-and-trace technologies be effectively implemented into the outbound logistics of a Swedish retailer in order to overcome individual challenges?

To answer the second and third research question and to fulfil the purpose, interviews with representatives of Swedish retailing companies will be conducted. The respective outcome will be reflected on a specific case of a Swedish sporting goods retailer, in order to develop a custom-made solution. Furthermore, the gathered data will be used to develop solution recommendations for each interview partner. It therefore shows the individuality and complexity of implementing a track-and-trace system within the outbound logistics of a cross functional operating company.

1.4 Scope & Delimitations

The scope of this Master thesis represents the possibilities of track-and-trace in the outbound logistic in general, and further practically applied to a Swedish retailer. Therefore, all commonly used tracking methods of the industry will be looked at and evaluated regarding their performance. The researchers are going to analyse the possibilities of each method and compare it to what is requested in its usability. This study only focuses on retailing companies in the Swedish market, because this thesis is written in cooperation with Intersport Sweden. Due to the specific track-and-trace problem in

6

relation to the logistics service provider, only other retailing companies who also use the same logistics service provider can be regarded. This study aims to investigate the processes of different Swedish retailing companies in order to understand their track-and-trace challenges and to apply their experiences to the main case example of this thesis. Companies from other countries can be excluded due to the unique processes and possible different hardware. In the practical example, the researchers are generating a custom-made solution for the Swedish retailer, where some aspects might be expandable to related companies. For some requirements, a less accurate but inexpensive and easy-to-install solution might be applicable. For other products, an accurate and worldwide real-time tracking might be the requirement. This said, the Master thesis hopes to untangle the currently available product portfolio and to give insights about the technologies available. It therefore is an analysis about the current technologies as well as an advisory how track-and-trace could help to overcome individual challenges of each regarded Swedish retailing company. It should also raise awareness about what kind of benefits a successfully implemented track-and-trace system can have for a company. Overall, the researchers are looking at the system from an usability and not from an IT perspective. Thus, this thesis comes along with certain delimitations. First, the thesis has a business background and is looking into IT details of the implementation only briefly. The business focus mainly considers a potential cost analysis for the main case example to effectively implement track-and-trace technologies in their outbound logistics. Furthermore, this study focuses more on the outbound logistics processes and less on the technical IT specifications for implementing track-and-trace systems. This thesis only investigates trace of outbound logistic and not the possibilities of track-and-trace in general. Additionally, the researchers only look at it as a business tool and do not take other applications of the technique into account. For example, track-and-trace is widely used in the observation of the behaviour of wild animals, as their movements can be tracked. The example of one specific Swedish retailer is looked at in detail, whereas the reality can be different for other companies. Lastly, the overall impact of COVID-19 must be acknowledged since it effected the overall conduction of the study negatively. 1.5 Outline

Within this section, an overview about the structure of the thesis is provided. The introduction gives a first overview about the background of the studied topics and

7

what challenges are present. The problem discussion follows respectively, which ends up with the purpose as well as the research questions of this thesis. Next, the scope as well as the delimitations of the thesis are presented. The second chapter covers the research methodology, where philosophy, approach as well as the strategy are going to be introduced. It also provides insights about how the data was collected, analysed and how the research quality was ensured. Chapter three looks at how track-and-trace is viewed within the academic literature. Therefore, a literature review about the possibilities of a variety of methods is presented. Empirical findings as well as the practical application for this business case example are provided in chapter four and five. The aim with the example is to show how a practical implementation could be conducted and what points are important to be considered. Chapter six completes the paper with a conclusion and a presentation of the managerial implications. It also provides further limitations as well as gaps which can be part of future research.

8 2. Literature Review

______________________________________________________________________

This chapter provides an overview of the literature in the research area of track-and-trace technologies in outbound logistics. Thus, a more comprehensive although general understanding of the topic of this thesis is given, so that it acts as a foundation for the empirical study.

______________________________________________________________________ 2.1 Definition of track-and-trace in logistics

Track-and-trace refers to IT-supported systems for determining the processing or delivery status of an object within a physical supply chain of a production or logistics company (Hassan, Ali, Aktas & Alkayid, 2015). According to Oliveira, Noguez, Costa, Barbosa and Prado (2013), the core task of a track-and-trace system is to create end-to-end transparency within a logistics chain so that customers, business partners and the logistics company itself can see the exact production or delivery status at any time. With reference to Shamsuzzoha et al. (2013), a track-and-trace system is essential for managing integrated logistics networks and improving customer services. In the literature, different authors define real-time tracking and tracing in different ways. According to Baresi et al. (2016), traceability has two characteristics: first, the attribute that makes it possible to determine the current location of a shipment, and second the registration and tracking of parts, processes and materials used in production by product identification numbers. Lin et al. (2013) define traceability as the ability to trace the whole supply chain processes backwards after delivering the materials and products. Furthermore, the wordings product tracking and product tracing have to be distinguished. According to Shamsuzzoha et al. (2013) product tracking refers to the location of a product due to its value and risk of loss. On the other hand, product tracing refers to the product and transportation history to identify the source of a quality issues. Furthermore, Främling and Nyman (2009) divide track-and-tracing into a forward and a backward part. Forward tracking is defined as the determination of the location of products along the supply chain process, whereas backward tracing refers to the identification of defective or lost articles in the logistics network. A track-and-trace system is not only limited to the level of manufacturers, but also to the entire supply chain networks. Supply chain networks can be seen as an integrative approach to coordinate the handling and control of materials and goods from

9

the origin to the end consumer (Baresi et al., 2016). For Shamsuzzoha et al. (2013) proper track-and-trace of all the necessary information from the supply networks is necessary for efficient and effective management. Thus, the respective supply chain partners have to collaborate closely together and define the track-and-trace requirements to enable a supply chain visibility. Klein et al. (2009) divide the respective stakeholders into two groups: suppliers and industrial/end customers. On the one hand, supplier and industrial customers require track-and-trace approaches from a business-to-business perspective. On the other hand, end-customers place track-and-trace requirements from a business-to-customer perspective, where e.g. the involved parties differ significantly

2.2 Technical foundations of track-and-trace systems in logistics

Track-and-trace systems can be implemented via terrestrial systems and satellite-based systems. Terrestrial systems are all systems that are based and implemented on earth, whereas satellite-based systems require corresponding hardware in space (Pavkovic, Berbakov, Ulianov & Hyde, 2016). According to Klein et al. (2009) the tracking of shipments is usually carried out via terrestrial systems. For this purpose, shipments (e.g. packages) are provided with machine-readable labels. Based on the order data, the destination or delivery date can be assigned to a shipment. Automatic sorting stations can read these labels and thus guide the shipment to its destination. Each time a scanner hardware reads a label or transponder, the information is sent to a central database. The data of the central database is prepared according to the requirements and made available to the respective interest groups via internet portal, ERP system or e-mail (Kothris, Beach, Allen & Karlsson, 2001). For the tracking of larger loading units, which also cover longer distances (e.g. trucks), mainly satellite-based systems such as GPS are used. A connection to the company headquarters is established via the mobile phone network and thus the exact position of a loading unit is displayed (Kothris et al., 2001). According to Deslandes, Tronc and Luc-Beylot (2010), the fundamental difference between terrestrial and satellite-based systems is the number of status messages for determining the position of objects. Terrestrial systems only issue status messages when individual defined process steps are completed. Such systems are referred to as process step-related or discrete track-and-trace. Discrete track-and-trace requires a line structure in which the objects to be tracked pass through individual process steps one after the other. Along this value-added chain, the exact location of an object can be determined for each step in the

10

dispatch or production process via e.g. RFID, barcode or 2D code (Kandel, Klumpp & Keusgen, 2011). In contrast, satellite-based systems identify objects continuously in real time, which is known as continuous track-and-trace systems (Kothris et al., 2001). 2.3 Basics of position determination

For the determination of the position of an object, the literature offers a variety of procedures. The variety of different systems of position determination can be derived from the number of different system requirements for a position determination system (Lee & Ferreira, 2002). Trilateration and triangulation are the main principles for the determination of a position. With an increasing distance, a transmitted signal loses its strength. A receiver realizes where each signal source is located in the area. If the receiver can receive signals from at least three sources, the signal strength makes it possible to determine the location of an object (Zhang, Xia, Yang, Yao & Zhao, 2010). According to Lee, Woo, Park and Kim (2014), trilateration is a technique that uses distance measurements to determine the respective position. This requires at least two known reference points on a two-dimensional map. Three known reference points are required for an accurate position determination. In addition, knowledge of the respective distances between a reference point and the searched point is essential. Lee et al. (2002) and Lee et al. (2014) mention that the distance can be revealed by means of a signal sent from the reference point to the searched point. For this purpose, the difference between the transmitting time and receiving time is determined. Afterwards, the distance is obtained by relating this to the signal speed. The searched point is therefore located on a circle around the reference point, which is called a standing line. The trilateration with a reference point does not result in a unique point. To improve the determination of the searched point, a second reference point is required. The result is two circular stand lines that intersect at two points. Logically, the desired point is located at one of these two intersections (Zhang et al., 2010). To determine which of the two points is the actual location, further information must be obtained. In practice it is often possible to exclude one of the two points. If this is not the case, a third reference point is required. The intersection of all three reference points is the position searched for (Lee et al., 2014). The second method for determining a position is triangulation (Lee et al., 2002). In contrast to trilateration, the angles are additionally used to calculate the position of an object. For a two-dimensional position determination, a length and two angles are required. The angle

11

is defined as the angle crosses the receiver. The delay of the signal indicates the angle and is called the "Time of Flight”. The “Time of flight” measures the time between the transmission of a signal between the reference point and an object (Zhang et al., 2010). According to Lee et al. (2014) this can be determined using special antennas with precise directivity. A distance measurement is not required for triangulation. As a result, straight lines rather than circles are generated here as standing lines. Thus, the determination point is at the intersection of the two standing lines.

2.4 Auto-ID systems in logistics

In order to gain a profound understanding about track and trace technologies, it is important to examine how they are currently embedded in the logistics and supply chain context. Auto-ID systems are considered as a subsystem of logistics systems. A logistics system is defined as the efficient transfer of cargoes or persons accompanied by the flow of information (Shamsuzzoha et al., 2013). A main characteristic is that the state of a cargo can be defined at any process time by an ordered pair of numbers. This requires an efficient system of transmitting information about the products or materials (Zajac, 2015). According to Shamsuzzoha et al. (2013), Auto-ID technologies enable the track-and-trace of the location of products in the supply chain. Lin et al. (2013) mention that in the supply chain, products or materials typically must travel within and across companies. Auto-ID technologies can be used to track and control the flow of the products between locations like warehouses, manufacturing and assembling lines. In an Auto-ID network, data about the same product are often generated at several nodes within the supply chain. To achieve full traceability of the product history these nodes must be connected and integrated by an overall system (Sun, 2017). According to Oliveira et al. (2013) and Do et al. (2006), this data integration is essentially influenced by several aspects. The Auto-ID system must be flexible in terms of constituting a dynamic environment, in which new nodes can be added to capture all product information. Do et al. (2006) refer to scalability because an Auto-ID node typically collects and manages a large amount of event data, which is further growing over time. Hence, the Auto-ID system must deliver corresponding optimization and analysis mechanisms in order to provide the required information at any time. Lastly, the authors mention security. Product information are often collected among others at the time of manufacturing, which can include sensitive detailed information for other companies. Therefore, it is necessary to address the inter-

12

and intra-company needs for data by providing corresponding views. According to Zajac (2015), Auto-ID systems are commonly used to address several issues in the supply chain context. The increasing widespread use of electronic data interchange and common databases which contribute to collaboration between clients and suppliers require interconnected systems and information exchange. Furthermore, Baresi et al. (2016) mention that companies are facing increasingly difficulties by providing the required information up the supply chain, due to growing production rates and customer requirements. Auto-ID systems enable companies to solve these problems by capturing information in real-time and connecting the databases (Zajac, 2015). Moreover, companies are facing increasing personnel costs due to growing requirements of added services within the supply chain. With a respective Auto-ID system processes can be optimized, and the impact of human mistakes and errors minimized (Gnimpieba et al., 2015). Hence, the integration of Auto-ID systems in the supply chain context can help companies to overcome the increasing challenges of supply chain visibility and information exchange through the respective channels.

2.5 Outdoor positioning technologies in logistics

To track-and-trace products and goods within logistics certain technologies exist. Regarding the outbound logistics of companies, the researchers analyse existing academic literature to examine which technologies are available. As already mentioned beforehand, outbound logistics is the interface between the internal and external environment of a company. Thus, outdoor and indoor positioning technologies that can be related to outbound logistics have to be considered.

2.5.1 GPS

As already introduced by a variety of papers, GPS tracking is a widely accepted and used form of outdoor tracking (Chirakkal, Myungchul & Dong Seog, 2014; Sultana, Tahsin, Reza & Hossam-E-Haider, 2016). In most articles, the literature focusses on the outdoor applicability and little research has been done regarding the practical application on an indoor application. This is mostly due to limiting factors like metal and walls, where the signal is reflected in several ways and therefore disturbs the direct sight to the satellites, which makes an accurate tracking challenging. In one indoor experiment, Lohnert, Bar, Gohler and Mollmer (2010) found out that one localization of a position consisted of three reflected paths and 417 refracted paths, where a majority was refracted more than

13

once. For an accurate localization, GPS needs at least four satellites to localize the GPS module on a map (Kandel et al., 2011). Piras and Cina (2010) discovered that the accuracy of localization software using GPS is improving and it is starting to allow for the determination of 3D coordinates even under difficult conditions. A median accuracy of 4,5-8 meters could be obtained. A GPS tracking device normally consists of three major components, in which the GPS black box is the central part of the system with a fixed interval transmission of data to a server. The GPS antenna is responsible for a strong receiving power to calculate the incoming signals. To power up the system, a battery is needed with which modern GPS modules can operate at least 72h. This easy and universal setup is important for the management of all involved parties in the supply chain, since the signal can be integrated easily into several different data management systems (Klumpp & Kandel, 2011). In general, there is a differentiation between “event-monitoring” tracking, where the location is only tracked near a reading station or is actively triggered, and continuous tracking methods, where the location can be detected any time. GPS might be able to close this gap, since its localization can be tracked anywhere and is not requiring a constant mobile connection, so that it reports its location once requested. This leads to a more transparent supply chain, which has benefits for the shipper in form of better planning and more detailed information for the customer. This is valid for most outdoor environments, but it lacks in its indoor tracking capabilities within a warehouse (Kandel et al., 2011). Xu, Chen, Xu and Ji (2015) conducted several experiments and found types of pseudolite architectures for the indoor usage. Several pseudolites are located at corners of the building to simulate satellite constellations, which are not reachable from indoor locations. A repeater simplifies the synchronization of the constellation with an outdoor antenna to collect the GPS signal. This method reaches an indoor localization with only minor modifications of commercial GPS receiver. The articles agree that GPS is widely used due to its universal application into various systems as well as the low costs of monitoring the shipments on a global scale. The different authors presented ways of using GPS in an indoor environment, but it becomes clear that without direct vision of the satellite tracking abilities of the system suffer in terms of accuracy (Lohnert et al., 2010).

14 2.5.2 GSM

Tracking via GSM is done with the help of mobile telecommunication technology. Therefore, the area needs to be actively covered by a GSM network in order to be able to use this tracking method. The covered area is divided into a number of cells, where each cell is served by its own base station. Several base stations are combined and controlled by a mobile service centre. This serves as the interface between the mobile network and the Public switching telephone network, which is the aggregate of different telephone networks of the world (Hussain & Kwak, 2009). GSM is the most widely deployed cellular telephony standard in the world, present in more than 220 countries with nearly 800 mobile operators worldwide. GSM will be used at least until 2021, because 3G and 4G do not have a comprehensive coverage yet. Therefore, GSM is still the current standard for mobile tracking (Tian, Denby, Ahriz, Roussel & Dreyfus, 2015). The fingerprinting location method can be transferred to GSM tracking, which will be introduced in greater detail in the W-LAN section. During the offline phase, a database of the environment is obtained. With this basis, real-time tracking can be done in the online phase by comparing the current results with archived data (Arif, Ahsan, Hasan & Bhattacharya, 2014). In their basic work Drane, Macnaughtan and Scott (1998) described three different physical architectures that could be used to position GSM mobiles: mobile-based, network-based and hybrid positioning. Mobile-based tracking is a form of self-positioning, where the position is calculated by the trackable device itself. Network-based positioning, also called remote positioning due to the remote calculation of the position, uses the signal of the mobile in order to position it on the map, as the distances to each of the sender can be measured. Hybrid positioning could be best described by combining aspects of self- as well as the remote positioning. One example could be the work of Shih-Hau, Jen-Chian, Hau-Ru and Tsung-Nan (2008), who compared GSM to FM tracking. One of their findings was that a hybrid solution for areas where the GSM connection is weak could increase the active tracking area once FM waves could be also used additionally. This results in a variety of possible hybrid tracking methods. A major limitation on positioning receivers is the phenomenon known as multipath, where a signal propagating from a transmitter to a receiver may travel via several paths including the direct, line of sight path as well reflected paths. Therefore, the recorded times can be misleading and cause errors in time measurements (Drane et al., 1998). Varshavsky, De Lara, Hightower, Lamarca and Otsason (2007) showed in their experiment that the

15

system could correctly identify the floor up to 60% and in the range of two floors up to 98% of the time, which show the high accuracy. In a more current work by Tian et al. (2015), a stable tracking system could be established in a challenging environment like railway and subway transfer stations, where multiple reflections must be considered. Localization errors could be corrected, and the accuracy of the system was significantly improved. Time and space constraints have been considered, so that local oddities could be transferred to the times measured. In the early work of Drane et al. (1998) it already became clear that an accuracy of around 100m would be achievable. Varshavsky et al. (2006) achieved a tracking using mobile phones with a median localization error of two to meters in an indoor environment. This also shows that indoor localization with GSM only became possible within the last years, since GSM macro-cell could expand their range significantly. In the example of Arif et al. (2014), an accuracy of over 88% using a simple experiment setup with a Nokia 3315 could be obtained. This shows that GSM tracking technology is capable of accurate tracking results even without using highly advanced hardware. Varshavsky et al. (2007) conducted their experiments using several mobile phones which did not significantly affect the accuracy of the tracking, so that the technical side of it only plays a diminishing role. Therefore, this system can be easily installed and applied to a variety of contexts, where real time tracking could be used without the need for a huge investment. In the experiments of Varshavsky et al. (2006), the GSM tracking results were only slightly less accurate than W-LAN tracking, but with the benefit of having the ability to track them everywhere within the range of mobile service and not only in predefined W-LAN zones. It showed that room level localization using GSM traces is possible and GSM phones could make applications like in-building navigation and in-building coordination possible. One important hindering factor according to Varshavsky et al. (2006) is the unwillingness of network operators to make signal strength information public. By not publishing these interfaces, network operators can limit the provisioning of location-based services to their own offerings. In a later work Varshavsky et al. (2007) analysed in detail the benefits of using GSM as the tracking method. There are recently developed, more accurate methods of indoor localisation on the market. Their drawback is that they require custom infrastructure for every area in which the localization should be performed. Therefore, these methods have not seen a significant development outside high-value application. In contrast, GSM uses existing infrastructure and

16

therefore limits needed investments into infrastructure. The accuracy of the system might be sufficient for the desired purpose, so that investments into more accurate tracking methods would not pay off economically (Shih-Hau et al., 2008).

2.6 Indoor positioning technologies in logistics

Regarding the outbound logistics of companies, the authors will take a deeper look in the literature to examine which technologies are available for indoor positioning and therefore able to close the trace gap. As a basis it can be stated that track-and-trace in the indoor context is mostly applicable to distribution centres, since goods are handled in these facilities before they are shipped out to the end consumer. This procedure is called indirect distribution, since the final consumer is not geographically close to the factory in most cases, so that intermediaries are helping to deliver the final product. Therefore, functions and roles of a distribution centre will be looked at in greater detail. This literature review will not take the discussions among researchers about the layout as well as the most economic feasible solution and the location into account. Instead it looks at the functions a distribution centre can obtain, where track-and-trace methods could be useful supporting tools. Chen (2001) simplified the role of distribution centres by referring to them as the connection point between supplying sources and the demand points of the end customer. Distribution centre can also act as a warehouse when goods are stored before they are shipped out, but a warehouse cannot act as a distribution centre. It is also the location where operation efficiency is enforced, since the processes have a wide impact on the overall economic performance (Parikh & Meller, 2008). In many distribution centres, items stored are picked and distributed in less-than-case-lot quantities, which means the picking process is labour intensive as well as repetitive (Liu, 1999). They are also the place where new theories in order fulfilment are tested, which have implications for the overall supply chain (Garcia, Seok Chang & Valverde, 2006). Therefore, the tracking of every outbound shipment within a warehouse is a complex operation. The following section tries to introduce a variety of technologies, which would be suitable for taking on the task to locate objects within a warehouse and therefore overcome the inaccuracies of the outdoor positioning technologies.

2.6.1 Barcode

First, the tracking technology of barcodes will be analysed. In general, there is an important distinction between 1D and 2D barcodes in the literature. Vartiainen, Kallonen

17

and Ikonen (2008) describe the 1D barcodes as a mix of vertical lines and spaces, where combinations of these two are called elements and represent different characters. The standardized 1D barcode consists of 13 numbers and is used worldwide for the marketing of retail goods. However, the usefulness in today’s world is limited by the small data capacity of the barcode, since a rising number of applications require longer barcodes to encode more relevant data. Nevertheless, barcodes are still an important factor in today’s retail world as its universally agreed layout is applicable to almost all commercial situations. In general, whenever a barcode is read, the status of the shipment can be changed to the next required action. Once this is done, the ID of the person who made the change and a time stamp are stored into a database, so that transparency is increased and each handling step is documented (Vartiainen et al., 2008). Regarding Kandel et al. (2011) barcode tracking is part of an event-monitoring since no notifications are sent automatically to a receiver. Nevertheless, barcode tracking can easily be included into active tracking methods, since with scanning the barcode, the shipment can be connected to the transportation vehicle it is loaded into. With this twist, a single barcode becomes part of consolidated shipments, where forms of active tracking might be applicable (Shuyi, Zhiqiang & Yongquan 2013). Shuyi et al. (2013) talk about the new generation of 2D barcodes (QR codes), which consist of a black and white matrix in which information are gathered. The information can be stored in vertical as well as horizontal directions. There are several sizes of barcodes available ranging from Micro QR codes to 177 x 177 modules with their maximum capacity of saving up to 7,089 numerals. The size also depends on the available space of the product, so that Micro QR codes are often seen on small electronic components. Sensitive information can be stored into secured QR codes, where the access is limited. Also, the possibility of frame QR codes exist, which gives companies the possibility to visually individualize their QR code to be easily recognized. This is mostly done in advertising and does not seem to be relevant for track-and-trace purposes (Rajesh, Waranalatha, Reddy & Supraja, 2018). Currently, coloured QR codes with more data storing opportunities are developed, but this technology is not ready for the universal mass application yet. In the case of Wang, Hung-Lin, Shou and Wang (2018), several barcodes for different processing steps have been used to track single items. Therefore, the usage of different barcodes allows to separate the information of a shipment completely to ensure data security. Nevertheless, the test also revealed that these barcodes should

18

be easily separable so that it is clearly visible which barcode is relevant for each processor of the shipment to keep the productivity high. Additionally, the readability of the 2D barcode is enhanced since it can be recognized correctly even if the barcode misses up to 50%. Vartiainen et al. (2008) conducted experiments, which concluded that barcode identification with a regular cell phone camera is applicable to situations where sufficient light is available. Also, the accessibility is important since a sufficient angle between the reader and the barcode is needed to enhance the readability. This goes in line with the conducted tests of Billo, David Porter, Mazumdar and Brown (2003), where seven independent variables of the bar code quality were tested. Only the x dimension (ideal width), the bar growth/reduction as well as the symbol contrast seemed to matter in the fast readability of these codes. Once taking these factors into account, barcodes can easily be read by a manual scanner or a fully automatic reading machine. One major benefit of barcodes is the easy setup for a barcode system with just a simple smartphone camera and access to the internet. Also, the creation of a unique QR code is simple and at low cost with the help of electronic generators. As the logo only consists of black and white, printing costs can be limited. With this setup, costs can be decreased and a tracking system with an acceptable security level can be implemented (Melgar & Melgar Santander, 2014). Another benefit is the universal application of barcodes as in the example of M´hand, Boulmakoul, Badir and Lbath (2019). In a challenging environment for localization technologies like within ferries, where the high density of metal makes the localization with advanced tracking technologies difficult, an easy arrangement might meet the requirements if no real-time localization is needed. Wang et al. (2018) point out that one downside of barcodes is that they are easily damaged as they are exposed to damages. Once the barcode is ripped or damaged, there is no way to scan or update the status of the product. He therefore recommends a plastic protective layer around it once it becomes obvious that the product is at risk of being damaged. In his example he calculated that even with a 1% rate of ripped barcodes, labelling could still be beneficial for the overall revenue generation. Once the barcode is scanned, all relevant information can be displayed immediately and no entries must be made in order to have access to them.

19 2.6.2 RFID

Radio frequency identification technology has become one of the most adopted track-and-trace technologies in logistics and manufacturing in the recent years. Unlike other technologies such as barcode systems, RFID identifies objects from a distance and does not require a direct line of sight. RFID tags include more applications than barcode systems and can measure additional data such as product type and even environmental factors such as acceleration, temperature and charge level (Want, 2006). Furthermore, RFID tags that are located in the same defined space can communicate with each other and recognize each other without manual assistance. RFID has become widely accepted in logistics and is one of the most promising applications when it comes to high precision tracking. (Lu, Wang, Zhao & Zhai, 2018). RFID is a wireless communication technology with which stationary or moving objects can be identified contactless using radio waves. The identification is contactless with the resonance or reflection principle by electromagnetic or magnetic (inductive) coupling. Depending on the RFID frequency (near field or far field) the coupling element consists of a dipole antenna or antenna coil (Anandhi, Anitha & Sureshkumar, 2019). RFID tags can be divided into three different systems - active, passive and semi-passive RFID tags (Choy & Ng, 2006; Xiao, Yu, Wu & Ni, 2007; Heidrich, Brenk, Essel, Schwarzer, Seemann, et al. 2010). Xiao et al. (2007) mention that passive RFID tags do not require batteries or maintenance. Passive RFID tags are inductively coupled, powered and received by an electromagnetic field generated by the reader. More precisely, data transmission is achieved by inductive coupling between the coil in the reader and the tiny col in the tag. Passive RFID tags store the energy they receive until there is enough energy for the tag to transmit data (Heidrich et al., 2010). A clear advantage of passive tags is that they are smaller, lighter, and less expensive than active tags. However, due to the inductive coupling they can only be read from short distance of up to 3 meters (Xiao et al., 2007). Passive RFID tags also have an indefinite operational life and due to their small size, they fit into a practical adhesive label (Choy et al., 2006, Lu et al., 2018). According to Heidrich et al. (2010) active RFID tags, unlike passive RFID tags, can send data in predefined time intervals because they have their own power supply (battery). This enables them to communicate data at a higher range of up to 100 meters. Anandhi et al. (2019) mention that active RFID tags are predestined for indoor localization. Active RFID tags offer an integrated sensor technology (position, acceleration, pressure, temperature, charge level) that enables

20

additional information transfer about the tagged object. The advantages of active RFID tags are characterized by their easy programmability and integration. However, the costs factor and installation effort are still downsides that must be considered. Cost effectiveness can only be reached by purchasing large amounts (Xiao et al., 2007). Semi passive tags have a power source which increases their working range and throughput. The main difference between active and semi passive tags is that semi passive tags still require a passive response from the RFID tag to the reader (Choy et al., 2006; Kiraly, Helfenbein, Kiraly, Kovacs & Balla, 2017).

2.6.3 W-LAN

Since GPS tracking is hardly applicable to complex indoor environments, alternative active tracking methods have to be considered which allow for real-time localization. W-LAN localization techniques in general utilize metrics of received radio signals. Traditional metrics are angle of arrival (where does the signal come from), received signal strength (how strong is the received signal), time of arrival and time difference of arrival. Time of arrival uses the absolute time it takes until the synchronous signal is received, whereas the time difference of arrival approach uses the time difference between sending and receiving the signal as the basis (Heidari & Pahlavan, 2008). Another important tracking technology for W-LAN is the discussed trilateration method. The distance from several W-LAN access points to the trackable device is measured (Torteeka, Xiu & Yang, 2014). The accuracy is improved with more access points in direct line-of-sight to which the trackable device can build up a connection. More access points are equivalent to more data about the physical distance to different routers which can be measured (Torteeka et al., 2014). Xiang, Song, Chen, Wang and Huang et al. (2004) on the other hand found out that after more than 100 measurements, the accuracy of their tracking system was not remarkably improved with further entries during the setup phase. This indicates that there is a threshold of how much data should be processed. The downside is that the accuracy of the method is easily affected by noise and the impact of closely located electronic devices, which is counterproductive for a tracking method in the context of a warehouse. The experiments were conducted in a direct line-of-sight to the receiver, which would be additionally challenging for this tracking method once this parameter of the test setup would be changed to the reality of a warehouse (Khan, Kai & Gul, 2017). Regarding Han and He

21

(2018) most of the current used indoor positioning methods are based on the received signal strength fingerprint recognition algorithm and are therefore called fingerprinting localization. This algorithm creates a database of the designated location area signal strength and compares it to the real-time collected signal strength. With this method, an accurate location estimation can be achieved. It does not require line-of-sight measurement to the sending nodes, which is beneficial in complex environments, where the line-of-sight can be easily blocked. In general, the system works well with unchanged environments, where influencing factors stay constant over a period of time. This technique consists of two phases. During the offline phase, a database or radio map of reference points is created, which serves as the basis for future localizations. During the online phase, the location would be obtained by using pattern-matching algorithms, which compare real time signal strengths with those recorded during the offline phase (Chirakkal et al., 2014). The overall fingerprint location technology is further divided into two categories. The first type would be the positioning method using an average of the signal strength for each access point, so that the user location can be estimated using a reasoning algorithm. The second type is a probability-based localization method, where conditional probability is used in order to estimate the user´s position (Han et al., 2018). Advanced forms of fingerprinting localization methods use context-aware information, such as a site plan, to reduce the special density of required wireless access points (D´Souza, Schoots & Ros, 2016). In some conducted experiments by Izquierdo, Ciurana, Barcelo, Paradells and Zola (2006), the time-of-arrival data was used to estimate a position. Time-of-arrival is part of the round-trip measurement and therefore a set of estimations with zero distance had to be obtained. An average measurement error of 0.64m regarding all cases could be achieved, when factors about the area have been considered. Therefore, W-LAN tracking is never an off-the-shelf solution but an individualized product. In the harder-to-locate environment of Jathe, Lutjen and Freitag (2019), an average location accuracy of 2,33m with the best possible algorithm combination could be achieved. The results of Xiang et al. (2004) are going into the same direction, which proofs that the development of an adjusted algorithm can significantly improve the results of the tracking system. Since the tracking accuracy was within 4m in 95% of all cases, this method of tracking is applicable for the practical commercial use if this range is enough for the purpose of the business (Xiang et al., 2004). In related papers, localization is done with the help of a grid, where vectors are pointing towards the place where the requested position can be localized. This

22

form of tracking is mostly useful in undistracted environments, where less distractions are taking place and therefore less suitable for track-and-trace in the context of logistic (Smallbon, Potie, D`Souza, Postula & Ros, 2015). Torteeka et al. (2014) conducted an experiment, where they combined the W-LAN trilateration technique with the fingerprint method. With this approach the positioning accuracy could be improved, the tracking system became more robust and the approximate positioning is continuous. Hence, this shows that once techniques are combined, a robust solution can be developed, which can mitigate the downsides of each tracking technique. In the work of Kirsch, Miesen and Vossiek (2014) as well as related papers regarding the Internet of Things, W-LAN is often used to transport information to a central server for further processing. This type of connection to a mobile phone or a different device cannot be seen as a tracking technique, since only information is exchanged and the method is not used for location purposes. In these setups, W-LAN is part of a hybrid localisation technique but not the source of localisation. One further example would be the wireless pick-by-light system of Asghar, Lutjen, Rohde, Lembke and Freitag (2018), where W-LAN is used to establish a more inexpensive form of a picking system with optimized routes. To conclude this section, it becomes clear that fingerprinting tracking as well as the trilateration method are the ones suitable for indoor tracking in a logistical context (Torteeka et al., 2014). Hybrid approaches are uprising ideas, which require more work to detect their full potential. 2.6.4 Ultra-Wide Band

Indoor positioning is often characterized by a high demand for precision and accuracy and can be affected by a variety of objects and signals. In recent years, Ultra-Wide Band (UWB) gained increased interest in indoor positioning, because it enables highly accurate positioning (Alarifi, Al-Salman, Alsaleh, Alnafessah & Al-Hadhrami, 2016). Zuin, Calzavara, Sgarbossa and Persona (2018) defined UWB as a radio frequency technology that spreads information out over a wide spectrum of radio frequencies. This enables UWB to transmit large variety of data while consuming only little energy (Alarifi et al., 2016). Promwong and Southisombat (2016) additionally mention that the time difference of arrival can be used with UWB in order to determine the distance between the reference point and the target. According to Pourhomayoun, Zhanpeng and Fowler (2012) UWB is defined as a baseband, impulse, and carrier-free technology. UWB radio communicates with high speed data rates in the respective area

23

by transmitting extremely short pulses of radio signals (Alarifi et al., 2016). Thus, the high bandwidth of UWB offers the possibility to transmit huge amount of data and the low frequency of UWB enables this technology to overcome the indoor positioning challenges of signal interferences through obstacles such as objects and walls. According to Pahlavan, Krishnamurthy and Beneat (1998) the high accuracy of UWB is very suitable for the tracking of different applications, such as mobile devices and humans in an indoor environment. The literature reveals different features of UWB positioning. Alarifi et al. (2016) state that UWB can also be used for the transmission of near-field data. As already mentioned, the high bandwidth and the extremely short pulses enable the signals to pass through obstacles and reduce the impact of signal interferences. This makes UWB a possible solution for indoor positioning in comparison to other technologies (Bastida-Castillo, Gomez-Carmona, De la Cruz Sanchez, Reche-Royo, Ibanez et al., 2019; Alarifi et al., 2016). Furthermore, UWB provides a high accuracy rate that enables positioning within three to seven centimetres. Therefore, UWB is mainly suitable for indoor locations that require a highly accurate positioning. In contrast to other positioning technologies such as W-LAN and barcodes, UWB transmits data over distance without requiring a direct line-of-sight. Furthermore, it is not affected by noises or other devices due to its high bandwidth of radio signals. A drawback of UWB are that the investment and installation costs for UWB are relatively high in comparison to other technologies. This is due to the fact, that UWB localisation requires at least three receivers to receive signal strength at any given time. These readers are expensive and must be precisely synchronized down to a nanosecond to accurately calculate the location. Moreover, the installation effort is increased, because to keep the readers synchronized, they are often connected by cables (Pourhomayoun et al., 2012). Thus, when considering Ultra-wideband as a track-and-trace solution, it should be carefully evaluated, whether a highly precise positioning is necessary due to the high investment costs.

2.6.5 Bluetooth Low Energy

Regarding Yang, Poellabauer, Mitra and Neubecker (2020) Bluetooth-based indoor localization is a long existing approach. Bluetooth classic was the used technology of position determination originally. However, due to its inefficiency this technology has not been widely used in the past, because the length of the connection process between the devices required too much time. With the development of the Bluetooth 4.0 (including

24

BLE) the situation changed significantly. BLE is characterized by a relatively low energy consumption and the configuration options enable a more efficient positioning in comparison to the previously mentioned W-LAN-positioning. According to Kriz, Maly and Kozel (2016) Bluetooth Low Energy is an emerging wireless technology for short-range communication that is designed as a low-power solution for control and monitoring applications. Bluetooth is an already widely established technology (e.g. in mobile phones, laptops, automobiles, etc.) and can thus benefit the implementation of BLE (Gomez, Oller & Paradells, 2012; Dalkilic, Cabuk, Arikan & Gurkan, 2017). BLE beacons can be considered as the most established BLE applications (Faragher & Harle, 2015). They are small devices that are used to illuminate the respective area by continuously broadcasting a signal to nearby BLE receivers. This enables devices such as smartphones or tablets to send or receive data packages when they are close to one or several beacons. The usage of BLE beacons is very suitable for indoor positioning and navigation of human in indoor environments (Faragher et al., 2015). Vasconcelos, Figueiredo, Almeida and Ferreira (2017) mention that the downsides of the adoption of BLE include the lack of support for large and dynamic data transmissions, security and privacy concerns and interoperability with other wireless technologies. According to Dalkilic et al. (2017), localization within buildings is often determined by W-LAN networks. Due to the often-complex structure of a building W-LAN signals can usually not cover all areas, which results in positioning inefficiencies. Kriz et al. (2016) agree and propose that these areas can be additionally covered by additional BLE beacons. BLE beacons are characterized by their relatively low price, small size and independence of an external power supply. Thus, they can be considered as a possible supplement to an existing W-LAN network. However, a high density of BLE beacons is needed to fully cover an area, which results in high investment costs (Yang et al., 2020). Hence, areas covered with weak W-LAN signal can be additionally illuminated by BLE beacons. 2.7 Implementation factors of outbound tracking technologies

The literature focusses on several factors, which are generally influential for the efficient performance of outbound logistical flows. One basic influencing factor is the price of the goods as well as the connected logistical costs. If these figures take up a big proportion of the total costs, a transparent supply chain becomes a key issue, which is worth investing in (Marques, Soares, Santos & Amorim, 2020). As pointed out by Shahmardan and

25

Sajadieh (2020) in their recent work, it is important to consider the mode of truck loading. A trade-off between waiting for full truckloads in comparison to sending out multiple trucks to one location is ubiquitous. In some circumstances, partial loading could be more desirably in comparison to waiting for complete loads to be reached. Hence, the quantity of shipments is of importance once outbound shipments should be tracked efficiently. Additionally, the characteristics of the tracked outbound shipment are highly relevant. One example given by Ling and Huang (2019) are the requirements for food shipments and the connected benefits of this excessive tracking. These products are not high in value but need extensive tracking to guarantee food safety. The same applies for critical safety equipment, where quality issues become a matter of life and death as experienced during the COVID-19 epidemy. Hassan et al. (2015) looked at several factors in more detail and identified six broad categories of hindering factors once analysing the implementation of Auto-ID systems. Besides the mentioned categories, technological issues play a key role. The connection to external partners as well as the own IT capabilities could be limiting factors. This is closely related to currently used tracking applications and technologies, given that they might be easily expandable. This also minimizes implementation costs, which make more sophisticated tracking solutions appealing to the top management. Another important outcome of their study was the role of management. Standardized shipment boxes are easier to track due to their almost identical product features. Decisions of the management can make a tracking system more tangible and therefore easier in its implementation. Also, the layout of the warehouse is of importance since some might be better suitable for implementing tracking technologies than others due to their structure. Moreover, compliance issues have been a key reason for implementing these technologies as it requires the involved parties to closely keep track of the goods once they move down the value chain. Van den Heuvel, De Langen, Van Donselaar and Fransoo (2014) analysed the effect of proximity on co-located logistical establishments. The most relevant advantage is the better usage of transport capacities. A close location to the logistical distribution centre is an investment itself due to higher location prices but reduces the need for excessive tracking of the outbound logistical flows because of a less challenging handover to the postal provider. Besides the factors mentioned, a variety of other influencing factors are relevant. Due to the individuality of each firm, influencing factors distinguish between each company. Hence, only the most general factors out of the literature have been mentioned.