DEGREE PROJECT IN PRODUCTION ENGINEERING AND MANAGMENT, SECOND CYCLE, 30 CREDITS
STOCKHOLM, SWEDEN 2018
Real Time Location System -
Industrial Implementation and Future Potential
An implementation project to evaluate different wireless technologies
and perform a cost analysis for a specific material handling process
within a global transmission manufacturer for heavy vehicles
FELIX WOLTERS
Abstract
This work is investigating an intralogistics material handling problem within a global manufacturer of heavy vehicles. Real-Time Location System solutions are investigated based on their wireless technology used for communication in an environment where many metallic semi-finished parts can interference with wireless signals. It is analyzed if the defined use case can be solved by an “off-the-shelf” solution without major infrastructural changes. In the end it is also discussed which prerequisites are required to implement RTLS solution in the context of other Industry 4.0 technology and development of production systems in the future.
sammanfattning
Table of content
1. Introduction ... 1 1.1 Background ... 1 1.2 Problem Description ... 1 1.3 Delimitations ... 2 2. Methodology ... 33. Current solutions established at Scania Transmission Manufacturing ... 4
4. Theoretical background ... 7
1.4 Auto identification (Auto-ID) ... 7
1.5 Real-time Location Systems (RTLS) ... 7
1.5.1 Radio Frequency Identification (RFID) based RTLS... 10
1.5.2 Bluetooth Low Energy (BLE) based RTLS ... 11
1.5.3 Ultra-wide band (UWB) based RTLS ... 11
1.5.4 5G based RTLS ... 11
1.5.5 Global Positioning System (GPS) based RTLS ... 11
1.5.6 Wireless local Area Network (WLAN) based RTLS ... 12
5. Results ... 13
1.6 Link Labs RTLS solution ... 17
1.7 Business case Link Labs ... 18
6. Evaluation ... 20
1.8 Cisco and ARISTA Flow solution... 20
1.9 Link Labs solution Air Finder ... 20
7. Discussion ... 22
8. Conclusion and future outlook ... 25
Table of Figures
Figure 1: Scania wooden pallets to handle semi-finished parts ... 4
Figure 2: Identification flag for material handling of pallets ... 5
Figure 3: Signal flag for material handling... 5
Figure 4: Prototype material flow ... 6
Figure 5: Example of different tags on the market ... 8
Figure 6: Example of access point to distribute wireless networks ... 8
Figure 7: schematic representation of RTLS components ... 9
Figure 8: abstract illustration of a RTLS ... 9
Figure 9: market ready product suppliers for RTLS solution and selection process ... 16
Figure 10: Link Labs Asset Tracking Architecture Source: Link Labs ... 17
Figure 11: Link Labs Use Case example similar to presented problem Source: Link Labs ... 18
1. Introduction
This chapter will outline the initial circumstances and needs for the specification of this work. The background is described, and the scope of this thesis work defined.
1.1 Background
The industry of heavy vehicle manufacturing is challenged by global ongoing changes like climate restrictions and rising customization [1]. This leads to new powertrain developments which suit for example alternative power sources and are in line with more strict emission limits [2], [3]. Therefore, major changes in manufacturing procedure such as higher number of product varieties and an increasing prototype production to support the development processes of all kind are required. For this reason, it is essential to enable controlling of current production flows and enable real-time decision making for Work-in progress (WIP). This will enable an efficient control of complex material handling processes and demand driven flexible production with low levels of stock.
The topic of Industry 4.0 is seen as a driver for innovation in manufacturing. The concepts aim to provide the mentioned functions and shift the production towards a smart and self-regulated environment to ensure competitiveness on globalising markets [4]. In many traditional industries it is still rare to experience implementation of well-connected machines and devices due to high investment costs and limited knowledge about how to get started with the new technologies within the existing production environment. Many strategy papers such as [5] recommend creating a clear vision for the topics of Industry 4.0 and get started with collecting first-hand experiences in small pilot projects. This work will provide an example for Real-Time Location system implementation which are named as technology within the concepts of Industry 4.0. It is investigated if a single use case at the material handling process is suitable for a first implementation of an RTLS solution. It is one of the pilot project which should built up knowledge about future production support system which will enable a more efficient and flexible production in the context of industry 4.0.
1.2 Problem Description
In the transmission production at the plant of Scania CV AB in Södertälje, Sweden, components for transmission systems for heavy duty vehicles are produced. Today many logistic tasks are carried out by manual supervision and high effort is put into controlling material handling. One highly important task is the production of prototypes for future transmissions and supporting the design process with manufacturing skills. In the current work routines challenges increase to handle the complex and changing material flows of these parts through the ordinary serial production flow. Limited production capacities and rising customer demand result in tight scheduling where unforeseen stops and loss of control can have high impact of the business success. One essential problem in this production unit is the absence of continuous information about the status and location of WIP pallets to ensure production as planned. Manual labour is required to perform simple task like counting pallets and localisation of WIP in the workshop to steer production. With the presented shift towards a smart production in mind the described problem is dedicated to get started with Industry 4.0.
the production unit at Scania. Innovative technologies and solutions are preferred to gain knowledge and collect incentive for continuous improvement and innovation.
1.3 Delimitations
This work does not imply to provide a ready-to-use system for further implementation. The focus is to show opportunities with these technologies for future production flow controls from the perspective of logistics in the transmission production of Scania CV AB in Södertälje and find the motivation for a larger implementation.
This work does also not include an in-depth analysis on which technologies are most suitable for a unique solution outside of this use case. The infrastructure boundaries and the limited time period effected the selection of technology solutions and suppliers.
The evaluation criteria of security issue are not investigated at this stage of the project due to limited in depth knowledge about security mechanism in wireless transmission technologies.
The focus in this work is on technologies that are named in direct connection to RTLS solutions. Therefore, barcodes and Near-field communication (NFC) are not considered.
2. Methodology
This work is performed as project work. First the current status of material handling in the related areas are described including the existing disadvantages. This information is gathered on the base of employees’ knowledge within Scania. Logistic developers as well as end users like operators and material handling planners are sources for mapping the current situation in the material handling process at Scania Transmission Manufacturing. The demands for a future solution are stated including their limits due to the environment and circumstances of this problem. In close contact with many different department restrictions and difficulties are highlighted in the process of implementation.
After this a literature research on Real Time Location System (RTLS) is performed to describe the boundaries of different technologies as well as their advantages. This will include discussions with former project manager of RTLS implementation, KTH researchers and experts at the Hannover Trade fair 2018.
Since this work is focusing on a short time implementation of a solution potential vendors and market ready products will be investigated as next step in strong connection to the previously found circumstances and boundaries of this particular use case. Based on the literature research and the circumstances of this case a solution is selected to perform a demonstrator and showcase the functionality of this solution. In the end a discussion is done about the RTLS solution presented if it can solve the presented use case.
3. Current solutions established at Scania Transmission Manufacturing
In the transmission production at the plant of Scania CV AB in Södertälje, Sweden, components for transmission systems for heavy duty vehicles are produced. Today many logistic tasks are carried out by manual supervision and high effort is put into controlling material handling.The 4 main production steps at the transmission department are soft machining, heat treatment, hard machining and shot peening. There are many different parts manufactured spread over 3 close situated buildings with a total indoor size of about 44,000 sqm.
Within the production steps the parts are handled in single units with articulated 5 axis manipulator and conveyor belts. Depending on the part number and manufacturing step several handlings are required. The area where single units are handled are considered as lines of soft machining and hard machining. The consumption into the line at soft machining is done from wooden pallets which are delivered by fork lift trucks to the line. Raw material blanks are manually loaded to conveyor belts which deliver the raw material into the first manufacturing step. Depending on the part number different manufacturing lines are chosen by the production planners. Several soft machining line and hard machining lines are built up at the moment and more to come in future due to the rising demand and limited capacities. The hardening process is performed by several furnaces of different type spread over the 3 mentioned buildings.
After each line the semi-finished parts are collected in standardized wooden pallets as shown in Figure 1. The line is provided with empty pallets by manually driven fork lifts. The loaded pallets are transported to a storing location selected by the fork lift driver to wait for the next manufacturing step.
There is no automatical, or digital identification method installed at this point. The material handling is organised by the planning of the production planners in communication with the material handling department which includes the fork lift drivers. The only identification of different part number is done with the help of visual flag (Figure 2) which have the imprints of the part number including the week number in which the order was started.
Each fork lift truck has a dedicated route it should follow to meet the demand for transport at all positions. The routing of the fork lift is design by the logistic development department. The so called slingors describe a fix sequence of stations the fork driver has to follow. In case of any shortages of material or empty pallets at one line the operator has the opportunity to signals this to the fork lift driver by rising a manual flag as shown in Figure 3. By this the dedicated routing can be short-term adapted to avoided down time due to missing materials.
One additional step is the quality control in the measurement rooms. In each building one measurement room is located to check a selection of produced part based on statistical process control. The measurement operations are booked into the material flow as production step and follow a strict time schedule. Only if the checks of the selected parts are accepted the whole order of parts (several pallets) can move to the next production step. Because of this reason it is likely to happen that high volumes of semi-finished parts are stored at different location to wait for clarification for further processing.
Currently the manufacturing step of heat treatment is considered as bottleneck due to its long down time for maintenance and functionality errors. In case of longer downtime, it is likely to happen that the storage areas exceed their full capacities and pallets of semi-finished parts waiting for further processing are stored in the aisles of the fork lift truck paths. In this unlikely event it is the task of the process planning engineers and fork lift drivers to keep track of the different storage locations and sequence of
orders manually. It is considered a huge effort to manually keep track of the WIP status at these situations. It involves frequent counting of parts and pallets by well-paid labour.
In the case of the prototype production small batch size of up to 50 parts are in one order. In addition to the in-house production steps mentioned above the prototypes can also be send to external suppliers for specialized machining and processing. This results in a more complex and not ordinary material flow of these parts (Figure 4).
The manufacturing of prototypes is planned by the production planners and are scheduled between ordinary serial production. Due to new material flow and unconventional transport ways it is challenging for the material handling to keep track on where prototype pallets are storage or in which status the order of prototype production is at each point in time. It happens that a measurement room is requesting a specific prototype order that is scheduled to be measured and several process planners are out in the workshop to search for the pallets. To get control about this complex material flow for prototype pallets a central hub in one of the building was established. The goal was to design a central controlling hub for the pallets in acceptance to longer travel ways and additional handling needs. This material flow design improved the material handling of prototypes. The logistic operator at the hub can track where orders were sent and monitors the incoming semi-finished goods at the hub. Due to short lead times it still can happen that prototype pallets are sent directly to next production steps without crossing the hub. In this case the hub gets no additional information where a specific order is in located at any point in time. In conclusion the material handling processes are complex and lack a method of continuous monitoring. Especially the prototype handling process are very complex and high financial consequences can occur when production is not performing as planned. Due to this reason the idea of implementing a real-time location system is investigated because it can provide continuous information about the location and flow direction of the material.
4. Theoretical background
This chapter will introduce different wireless technologies which are suitable for RTLS Solutions. This literature research will serve as base for a test system selection for the industry problem.
1.4 Auto identification (Auto-ID)
Auto identification (Auto-ID) technology is used to automatically identify entities with the help of tokens with individual code. This information is collected in a computer system to support controlling functions like localisation of entities or recording of status information [7]. The most known technology for this purpose is Radio Frequency Identification (RFID) but also other technologies such as Near Field Communication or use shape recognition are used in several use cases. Which technology is best to use is highly depending on the circumstance of the use case and therefore requires intensive evaluation before implementing [7] [8]. One of the key function in future Auto-ID technology is the integration into a wireless communication network according to [7]. This enables the management of different tokens according to the physical unit they are attached to. Auto-ID technology is purely only related to a digital identification of physical entities. Other systems such as Real-Time location systems (RTLS) are built on top of auto-ID technology. As stated in [9] Auto-ID technology gives opportunities to make WIP visible and enable early decision making and reduce malfunctions in production. The functionality of localisation makes the RTLS the next step after Auto-ID solutions [7]. Cost for tags and the gained benefits decision driver between Auto-ID solution and RTLS. Since this problem description is focusing on a specific use case where localisation of entities is important the focus lies in RTLS solutions. Nevertheless, it is important to keep in mind that an Auto-ID solution can have a high benefit for the ordinary material handling process and have similarities with RTLS. It is also described that an Auto-ID solution should be a preparation step before an RTLS solution is established.
1.5 Real-time Location Systems (RTLS)
This technology focus on the localisation of tokens with the help of wireless technology. [8] and [10] understands RTLS as a technology that requires Auto-ID technology and is built upon it. RTLS are solutions to continuously monitor entities within a defined landscape [11]. In 1998 the first system was designed to monitor asset and people in a defined indoor area [11]. Where time critical decisions are needed RTLS provide additional information about the location of assets and people to quickly find them in space. There exist different technology groups which are used in RTLS systems. Radio Frequency (RF), acoustic solution, and visible light solutions. In this work RF solution are in focus due to the highly dynamic and harsh environment in industrial manufacturing. RF systems use WIFI, RFID, UWB, 5G, GPS and BLE technologies depending on the use case. RTLS Solution are separated into software and hardware components.
Therefore, passive RFID tags are most of the time very slim and small whereas Wi-Fi tags are bigger due to higher energy consumption.
The localisation sensors are either access points (Figure 6) or reference point distributed in the area to support the localisation engine. They distribute the wireless network and communicate with the tags. For this reason, most of them need electricity support and a fixture to mount them to wall, pillars and ceilings. Hardware wise there are also existing three different approaches. The first and most common
one is to use access points as localisation sensors. Their location is fixed and during the set-up time evaluated. The localisation engine is that calculating based on the different signal properties of a tag to each different access point the localisation of the tag. For this approach the access points have to be distribute very close to each other because the tags need to be in the range of multiples to gain a suitable localisation information.
The second approach is mostly used with passive RFID tag. The localisation sensors are used as choke points. Each time the access point is recognising a tag the information is generated that this tag is nearby in the range of this specific access point.
The third approach is to establish a decentralized localisation engine. The localisation is calculated inside the tag. Based on signal from battery powered reference point which are easy to deploy the tag can calculate it location. This information is then sent out through for example the established Wi-Fi infrastructure. With this approach the data traffic for given infrastructure is reduced and the level of
Figure 5: Example of different tags on the market
On the software side the localisation engine can calculate either based on the signal properties or with the help of localisation algorithms the position of the tag in a defined base. The middleware is connected to the evaluated locations and provides this information to different application such as dashboards or automated decision-making tools.
Figure 7 and Figure 8 show a schematic representation of an RTLS solution 1.tag, 2. localisation sensors 3. localisation engine 4. middle /application.
The software solution for localisation engine and middleware can be adapted to most of the mentioned wireless technologies. The key function of a RTLS solution is the positioning technology. To understand the different methodologies on how the location of a signal can be calculate the following will introduce different signal properties and positioning algorithms that are used in RTLS solution. When talking about position system two basic levels of gain a position are required. On the one hand the properties of received signal of any kind. [12] names four main properties that are used for indoor positioning systems.
Angle of Arrival (AOA) described the angle in which a signal is received by a receiver. Time of Arrival (TOA) or also called Time of Flight (TOF) is the property of the signal which described how long the signal travels from sender to receiver. Time Difference of Arrival (TDOA) is built on the idea of TOA but describes the difference in Time of Flight to different located receivers. The last mentioned is Received Signal Strength Indication (RSSI). It describes the power level of the incoming signal which correlates with the distance between sender and receiver. [13]. All these signal properties have different advantages and disadvantages for specific positioning use cases. For indoor position solution TDOA is stated with the highest accuracy whereas it also required expensive system designed to measure this signal property [12]. RSSI on the other is expected with cheap installation but lacks the accuracy benefit [13]. [12] well describes this evaluation in detail and also explains that therefore the positioning algorithms are essential to gain the location based on different received signal properties. The positioning algorithms filters the signals and translate them through calculation into an actual position
Figure 7: schematic representation of RTLS components
information. All this is happening in the localisation engine and is then forwarded to the middleware of the RTLS solution.
Triangulation is based on the AOA property and calculates the position for a signal based on its angle of single income [13]. Its advantage is a low-cost hardware structure and simple calculations for short range positioning. But the accuracy according to [12] is not as high as other algorithms. Trilateration for example is based on TOA and TDOA. High accuracies can be achieved but the system needs more complex and therefore also expensive calculations. Proximity is a relatively simple method that is based on the closest available receiver. Based on the RSSI property the position is estimated closed to the receiver with the highest strength. The downside with this method is that due to signal interference with the environment inaccuracy can occur. One method with a potential for a very high accuracy is fingerprinting. This method is based on a database of RSSI signal collected during a calibration phase at the start of implementation. The online received signal is then compared to the database values and based on that the position of the signal transmitter can be given. The relevant disadvantage her is that any change in the environment will require a new calibration of the database recordings have to be a perceives as possible [13].
The mentioned RF technologies will now be described including their strength and limitations. Important to mention here is that the difference in wireless technologies only includes the hardware component of an RTLS.
1.5.1 Radio Frequency Identification (RFID) based RTLS
The first patent within RFID technology was published in 1973 and the early development was started back in 1945 [14]. Back then mostly use cases in transportation and healthcare were identified. The RFID tag include a limited read and/or write storage for data. Depending on the type the tag the size can differ. Nowadays RFID technology can be divided in passive and active tag technology [15]. Active RFID technologies use a battery powered tag whereas passive RFID technologies use tags which receive the required energy through the signals send from a high-power reader (localisation sensor). [15]. Through this low frequency transmitted energy the passive tag is able to transmit a message to the reader. Low distance between reader and tag (0.5m – 2m) and low cost for tags ( a few cents) make this technology suitable for inventory tracking systems according to literature [15] [16].
Active RFID technologies can transmit signal without triggered by the environment due to their integrated battery. Long distances (90m) [16] between reader and tags are therefore possible but also the lifetime is limited to their battery capacity and varies depending on the size of battery and amount and frequency of data transmission.
RFID tags is very low compared to other solutions. [18]. Many references state that the price tag is critical to the wide spread use of RFID technology in logistic processes [19]. It is also said that for evaluating this technology the usage has to be seen in a bigger context to grant all the potential benefit that can result out of the new knowledge generation [19]. [20] states that RFID signals are highly influence by metal and fluids surrounded by them. Therefore, for a RTLS solution with RFID technology it is essential to test it in the given environment and find solution to reduce the signal interferences with the surrounding.
1.5.2 Bluetooth Low Energy (BLE) based RTLS
This technology is similar to active RFID technology and is used for short-range wireless communication [12]. Since the 4th version of Bluetooth the Bluetooth Low energy also called smart Bluetooth is available [21] [22]. It was developed by the Bluetooth Special Interest Group which oversees the standards and licence the usage of this technology [23]. Literature grants BLE high potential for future localisation solutions due to its standardisation, low energy consumption, low cost and high precision. [24] [22]. BLE tags are also battery power and enable longer distances than passive RFID technology but shorter than active RFID. One mentioned advantage of BLE technology is that it is integrated in many mobile devices by default. Mobile phones and computers have the hardware ability to communicated through the Bluetooth standard. Bluetooth operate on the Industrial Scientific Medical (ISM) band 2.4 GHz. [25]. Mesh solutions and even battery free tags are current topic in research. [25] [26]. [22] calls BLE a revolutionary technology in the field of communication in context to Internet of Things applications.
1.5.3 Ultra-wide band (UWB) based RTLS
Ultra-wide-band (UWB) is a high speed short range wireless communication technology with high data rate capabilities [12] [27]. A detailed UWB definition can be found at [28]. Since 2002 the band between 3.1 and 10.6 GHz is reserved for UWB indoor communication systems by the Federal Communication Commission [27]. UWB is situated with a high potential for RTLS solution due to its high precision capabilities [29]. This technology is more expensive in deployment compared to the other here presented technologies but due to is high penetration power and rising research it is to expect that this technology will play a bigger role in indoor communication technologies [12]. [30] states that further investigations outside the lab environment and inside the shop floor characteristics have to be performed to evaluate it potential for the future. [31] give an overview of current ongoing research and details about different standardisation of this technology. In conclusion this technology is of high interest in future but is still at an early stage of research related to RTLS solutions.
1.5.4 5G based RTLS
5G is the fifth generation of mobile cellular technology in the current state of development. [32]. This technology is based on the IEEE 802.11ac standard and is seen as enabler for Industry 4.0 solution in the wide field of industries. [33]. 5G is capable of data rates up to 10 Gb/s and will be market ready in 2020. [33]. It is said to change the wireless communication industries due to its rich data rate capability and high level of standardisation. [32]. In terms of RTLS solution the usage of cellular technology is sceptical. [34] states the accuracy of this technology for RTLS at about 50m -200m and recommend base station for indoor position services. It is to be questions if with the development of 5G new properties will increase the accuracy matter. The most named application of this technology is for critical communication network within the concepts of Industry 4.0 such as real-time monitoring of censored systems. [35]. This technology is at a very early stage of development and therefore no market-ready solution for RTLS exist.
1.5.5 Global Positioning System (GPS) based RTLS
indoor application due to the limited ability to penetrate concrete buildings and roof structure of buildings. [37]. The success story of the GPS for outdoor localisation give according to [38] the motivation to establish technical solution for indoor localisation. He mentions that there are many applications in industry and outside where localisation can give insights about customer behaviour process dynamics and other value information. The literature clearly states that GPS is not suitable for indoor positioning solution.
1.5.6 Wireless local Area Network (WLAN) based RTLS
5. Results
As the literature research has shown many different technologies for wireless communication and therefore also for RTLS solutions exist. The most named challenges for these technologies are cost of deployment in contrast to the gain benefits. Technically the interference of these technologies in a manufacturing environment is the biggest concern in terms of localisation.
Due to this conclusion an investigation was performed about market ready RTLS solutions which are easy to integrate into the existing IT infrastructure and consist of a hybrid solution which enables scalability of the system. The chosen demonstrator will focus on the initial problem description of localisating the high value pallets on the shop floor but the evaluation of the solution will be done in context of the bigger picture in future and other use cases. This is due to the fact that RTLS solution must deployed on a wide area and have therefore high cost accounts. Important to keep in mind for the prototype use case is that the flow of pallets is very dynamic. There is a high variety of paths a pallet can take through the production. Because of this reason an RTLS solution is required which give the relative location of the pallets in a coordinate system because it is not guaranteed that the pallets will pass given choke points during the complete handling process.
The market research about potential RTLS providers have unveiled that a broad number of suppliers claim to provide solutions. To suggest the suitable provider for this engineering problem the vendors were grouped into three different categories. There exist basic hardware providers, which are producing tag and transmitter hardware for the mentioned different wireless technologies. The second group are solution suppliers which use the hardware of the first group to design services which provide the demanded functionality. The last group are the Industry 4.0 infrastructure provider. They work either together with solution providers or directly with hardware providers to use existing infrastructure also for other purposed beyond the field of RTLS solutions. It is important to clarify these different levels of providers depending on the use case different types of vendors are suitable. In case of short term solution there is no time to develop solution high up from the rough hardware components. Ready solution with easy integration capabilities are required to bring value as fast as possible. On the other hand, when time is not high restricted the development of tailor-made solution within internal IT departments can be beneficial when enough use cases are present. But it has to be kept in mind that developing such support functions for production may not be the core value of a heavy vehicle manufacturer.
In a first step the suppliers were selected which provide their solution in the production industry. Many RTLS solutions are designed for healthcare application such as patient tracking or hospital asset monitoring. In this industry the environments are cleaner and well defined. The number of potential interferences is reduced compared to industrial production applications where harsh environments are present and heavy machinery can interfere with wireless technologies.
The second step selected vendors based on their primary wireless technology they use in their solution. Due to the fact that UWB as mentioned provides a high precision coming at a very high cost of this solution this was not the technology of choice. Since the present problem is very limited to a low number of parts the cost of a solution should be very low. As pointed out by the project directive it is also not require to have that high level of accuracy for the described problem. The fact that Scania is running another project with a RTLS solution based on UWB for AGVs installation drove this decision as well to not select UWB technology for this demonstrator. It was decided that a new wireless technology should be tested so there can be made comparison after the demonstration. Both the research has and an internal discussion within Scania has shown that passive RFID technology is connected to a large infrastructural installation. This would increase the cost for a solution dramatically and would go beyond the limits of solving the track and trace problem for the prototypes. Another factor for this decision was that a passive RFID tracking solution is already existing in one of the warehouses within Scania. There is existing already a strong knowledge base about the advantages and disadvantages of passive RFID technology and since this project should present new approaches passive RFID technology was not considered any more. The most favorited wireless technology in terms of precision for cost trade-off is BLE [13]. The literature research has shown that this technology involves a high level of standardisation which enable interoperability and is driven by a strong interest group which is even developing battery free tags. Many market ready solutions are using BLE for wireless communication and localisation on the shop floor. Additionally, there is not pre-existing knowledge about this technology as point of now. Therefore, testing this technology for the purpose of RTLS solution would generate new knowledge within Scania and would complete the wireless technologies that are suitable for localisation. As a result, solution providers with market ready technology based on BLE where selected for a more detailed evaluation. With this decision a new wireless technology can be tested on the shop floor with all the given possible interference and knowledge can be generated for further project.
its capability comes at a suitable price. The following chapter will describe the technical solution of Link Labs in detail and introduce the business case for the prototype use case.
1.6 Link Labs RTLS solution
For the demonstration after this work the solution from the American company called Link Labs was selected. Link Labs was founded in 2014 and is based in Annapolis, Maryland. The founders worked in projects for the U.S. Department of Defence, telecommunications industry and Intelligence Community and studied at the John Hopkins University Applied Physics Lab. They provide asset tracking and monitoring technology. Their RTLS solution is calls Air Finder Asset Tracking and is based on BLE technology. The system consists of three hardware components (Figure 10).
The BLE asset tags, BLE access point and BLE Reference points. The asset tags are battery powered and attached to the assets which should be tracked. They are equipped with an internal processor which is able to calculate its position based on the nearby reference points. The reference points are fixed and
installed at known geolocation points. During the set-up period this information have to be fed manually to the system. For precis localisation the asset tags are using the geolocation information of the reference points. It is also possible to gain a more inaccurate location information based on the connectivity to the access points. This functionality allows the user to scale the system depending on the specific need in certain area. It is possible to have a critical area tracked down to several centimetres with the help of multiple reference points whereas an open wide area can only be tracked by a choke point principle where the connection to the nearest access point is sufficient. The asset tags are designed to only calculate their position in case of certain events. Either a frequent calculation is made based on scheduling or an accelerator measurement initializes a new location calculation when the tag is moved. When the asset tag has calculated it position it then sent out the to the nearest access point. From here the information can be directed to the business intelligence solution either a cloud-based solution can be used to display the different locations or an integration into a local Manufacturing Execution System (MES) can be established. It is also possible to use existing WIFI infrastructure to transfer the gained information about the location of the tags. In Figure 11 a use case example provided by Link Labs is illustrated.
1.7 Business case Link Labs
As described in 3 the current material handling and planning is done mainly manually. This means a high paid engineer is often required to walk through all production facilities and find the required WIP pallets and identify their current status. An estimation was made that by introducing this Link Labs solution just to the prototype handling processes could save a third engineering position. This would give room for more development of material handling process and drive the continued improvement process.
The tags come at a cost of $16 per piece at an ordering amount from 100-499 tags. The access point cost about 200$ and can cover an area of 185 m2 each. The price for the reference point is about 17$ also
depending on how many are needed. Making a rough calculation with this pricing a total hardware cost for a 44,000m2 will end up at 50,050$. Additionally, an annual fee of 15,400$ to get access to the software solution and location calculation with the Link Labs interface for the number of up to 500 tags is required. Missing costs are here the installation of the 230 Access points and the overall set up costs. In the evaluation of this project these costs were neglected due to the reason that for any kind of solution
these costs will come into account. The variability in this cost according to the number of access points to install is considered as low. Therefore, the focus for comparisons are the hardware costs and the annual fees for maintaining the functionality of this solution.
Apart from simple labour cost savings other difficult to measure benefits result from the installation. The understanding of material flow principles will support the continuous improvement process. When such a system is installed in the complete area of the transmission department also other assets apart from the prototype pallets can be tracked. It is a new type of information generation about the processes
Figure 11: Link Labs Use Case example similar to presented problem Source: Link Labs
Nevertheless, the cost structure of RTLS system was considered as a major topic during this project. RTLS solutions are depending on a strong hardware infrastructure. Without this infrastructure there is no possibility to establish a reliable and cost-efficient function of localisation. In this particular project there is only a poor Wireless Local Area Network available which is due to high power consumption and low accuracy not suitable for localisation functions. After internal discussion within Scania and KTH as well as expertise talks at the Hannover Fair 2018 in Germany with RTLS suppliers the conclusion was made that RTLS are very costly to deploy when no basic infrastructure like electricity and Wi-Fi is available. RTLS Hardware cost depending on the wireless technology that is used and can vary. Overall many papers have stated that BLE technology results in the lowest cost. The annual cost of maintaining tags and software support for the localisation engine can vary for different suppliers. This means for this particular business case that as long as the basic infrastructural changes are accounted only to the additional functionality of localisation none of the market ready solution will be cost efficient. But when the cost for basic infrastructure change like electricity socket at the ceiling and Wi-Fi connectivity in all areas are split among other use case RTLS solution deployment can become cost efficient. In particular BLE technology promises to have a very low upfront hardware cost compared to UWB or RFID.
6. Evaluation
This chapter will evaluate shortly the Cisco solution and explain why their solution was investigated outside of this work and will highlight the advantages the Link Labs solution named Air Finder can bring to the current material handling process at the transmission department as well as describing the advantages outside of the present prototype problem.
1.8 Cisco and ARISTA Flow solution
One solution of high interest is the RTLS from Cisco and ARISTA Flow. Their system is built on top of existing WIFI infrastructure. They are using WIFI tags to locate asset in a defined area. Their solution was not selected for a demonstrator due to the reason that it was stated that WIFI signals are not well suitable for localisation purposed and heavy signal interferences can occur with rising number of tags. Another reason is the a relatively high tag price for the ARISTA Flow tags of about 110$ per piece as well as the complex set up of the system. It is required to have a trained engineer to install their solution and this would have exceeded the resources of this master thesis work. All the reasons drive the decision to not take their solution for a demonstrator in short term because it will not give the required fast first-hand experience as described. Nevertheless, Cisco’s solution could be a long-term solution for the transmission department and Scania due existing Cisco infrastructure and scalability of their solution. Therefore, a continuous discussion with the Cisco project team was established and an evaluation for their solution will continuous outside of this work. This topic will be picked up in the discussion section and future outlook.
1.9 Link Labs solution Air Finder
the technical prove of their solution in the given environment a decision can be made about the complete roll out. The technical capability is the second critical evaluation criteria. When talking about technical capability it is to define first which accuracy is needed exactly to solve the case. In case of the prototype problem there are different area where different levels of accuracy might be required. The Link Labs solution fits perfectly this purpose. In areas where high precision down to meters is required more reference points can be deployed whereas an open area where a rough localisation is sufficient no reference points will be deployed, and the localisation will only be based on the access point connectivity. Other solution can only provide one level of precision to the whole area. It could be a solution to enable in the whole area with a higher precision than even needed but the high precision comes at a higher cost which would as described exceed the resource of the prototype problem. Another technical advantage of Air Finder is the management of battery life of each tag. Computations are only performed when the localisation status is changed. In this matter the wireless communication is reduced to the minimum needed. In case of high number of tags to track, it can occur that the data traffic for the present wireless network is too high and will decrease in performance.
Additional arguments for testing a cost effective and easy to install system are the lessons learned that can be generated from this demonstrator. The handling of tags, how to attach, how to store them and maintain them are important factors that cannot be estimated without first-hand experience. This demonstrator can give insight into issue that can occur during ramp up of such a system. Another big topic that needed to be discussed with the supplier company is about who will own the data. All these topics are generally applicable independent on which supplier is chosen but they are very important to know and also to ask the right questions when a bigger roll out project is started.
These including the main reasons named before are why the Air Finder tracking solution was selected for a demonstrator. Because it is a new approach on achieving scalability of accuracy in different areas and it is easy to deploy. The cost for testing are low and a first cost calculation makes it suitable for a bigger roll up with the potential to solve the prototype material handling but first the technical capability and handling of the system is tested in the demonstrator. An important note here is that it must be questioned if the cost for installing their solution for all area where the prototype pallets can move to is cost efficient. The cost of deploying the proprietary access point will add a very high cost to this project which is not sure to be compensated by operational benefits. In general, the evaluation of the business case has resulted in the opinion that a deployment of RTLS solution for a relatively small number of pallet (200) in a very wide area (44,000 m2) is not economical because infrastructural changes are to
7. Discussion
The discussion on this work has to be seen from two perspectives. On the one hand the short time problem of handling prototype pallets in the facilitates of the transmission department in Scania and on the other hand the overall strategy towards smart factories and industry 4.0.
When considering the scope of this project limited to the prototype material handling problem it is hard to find a 100% solution which will solve the problems at a suitable price with no infrastructure changes and a small impact on the running production. Since a localisation function can only be provided for an area that is deployed with the right hardware it will affect all areas of current production. The solution which is working with the current infrastructure of Cisco is not that easy to demonstrater at low costs. Nevertheless, the problem is existing now and need to be solved because the material handling process will become more complex and difficult. A solution has to be established as soon as possible without doubt. This work has shown that there are existing systems that can provide this functionality with market ready products. But the cost and effort are too high to justify a solution only for a single use case. Because of that the implementation of an RTLS solution must be considered in a wider scope and maybe other use case where this technology can be used. To accelerate this process, it is recommended to include the potential end user at an early stage into the development and implementation process. For this reason, the demonstrator of the Link Labs solution should be performed with the focus of showing the end users the benefit and create a demand from their side. The demonstrator will lead to first-hand experience which is not present as state of now. This would mean that the demonstrator is not performed to solve the prototype material handling problem directly, but it will enable a fast development of an Industry 4.0 technology on the shop floor which is well designed in the need of the operators and material handler or other end users in production. The demonstrator for example could highlight problems in the handling of the tags which will affect existing work routines. There are many pitfalls which can result in a more ineffective handling process than the current stage when additional tagging routines are added. For example, tags must be place at reachable location on the pallets. There need to be storing units established and operators have to be trained on how to use them. All these issues can be experienced with a relatively low-cost demonstrator from Link Labs solution.
Another essential topic is the IT infrastructure development on the shop floor. This project has shown that the current infrastructure is not suitable for upcoming topic like RTLS in the field of Industry 4.0. The cost of preparing the wireless networks and infrastructure for the new functionalities is high and is slowing down the decision to bring new technologies on to the shop floor.
Additionally, this project has unveiled the missing communication between different function within Scania. There are many projects running in the field of Industry 4.0 and even several RTLS investigation are done in parallel without a central coordination of these works. This have led to time intensive investigation because information is not stored centralized in a structure order.
For this purpose, a strong and reliable network have to be established to perform all tasks that will support the core value creation process, the production of heavy vehicles. From the simple internet connection for personal computers and handhelds in the workshop over machine communication up till the information distribution for historical data analytics. These are all functions that are enabled through a strong IT infrastructure.
When focusing on the shop floor network infrastructure the literature research has shown that wireless networks will play a major role in future production. But as of now there is no strategic decision which wireless network is most beneficial for the purposes of Scania transmission. A suggestion would be to investigate all possible use cases which can benefit from wireless network technology and run a project to improve the current infrastructure to enable further Industry 4.0 development. A similarity can be seen in the availability of high pressure air supply for machining purposes. When only one purpose is considered which can benefit from an advanced high-pressure air supply it will be never considered as cost-effective to change the infrastructure. But when all machines and other functions taking into account the cost opponents a high value add which make it worthy to make change.
Here is also where the second approach is important. With demonstrators and test kits the end user can more easily define it needs and user stories can be described in a more detail way which will lead to a more tailored made development. Therefore, the demonstration of the Link Labs solution will give a more detailed understanding the demand on such a system and enable user different development. The experience and lesson learned from the demonstrator can then be used in the project to change the infrastructure. It is also important to have prior-knowledge in these technologies to request the right technical solutions.
This was also the intention why the Cisco solution was evaluated outside of this work. Cisco can be a key player in an infrastructural change to enable Industry 4.0 technologies. Their advantage of having already deployed infrastructure can be used as cost advantage. This will enable the required scalability where stepwise functionalities are built up. In a first step it could be demonstrated which functionalities could be enabled with the current infrastructure deployed. This would create a short-term benefit at a low cost and give some time and understand on all sides to start the bigger infrastructure project. For example, the functionality off Auto-ID technology can be partly integrated passed on passive RFID technology for the ordinary flow of material to start introducing a digital identity to the units of the shop floor. Other functionalities like maintenance sensor solutions or wireless analytic instruments are single use case which require a basic infrastructure.
But Cisco is also only one potential key partner due to its existing business relations with Scania. Companies like Robert Bosch or Siemens have developed a wide portfolio of Industry 4.0 technologies which can be purchased modularise in one whole system. RTLS is for example based on BLE technology at the exceed solution from Bosch. Siemens provides a RTLS solution that is using additionally low powered screens on their tags to make the operator interface easier. Another important source for potential partner in this topic could be the network within the Volkswagen Group there are maybe already existing strategic partner and investigation made.
8. Conclusion and future outlook
9. Bibliography
[1] M. W. Rueger, F. Vernersson and C. Brash, "The truck industry in 2020 - How to move in moving markets," PwC, Berlin, 2014.
[2] B. Sharpe, N. Clark and D. and Lowell, "Trailer Technologies for Increased Heavy-Duty Vehicle Efficiency: Technical,," SAE Int. J. Commer. Veh. 7(1), pp. 238-261, 2014.
[3] J. Lowrie, H. Pang and G. Ngaile, "Weight reduction of heavy-duty truck components through hollowgeometry and intensive quenching," Journal of Manufacturing Processes 28, pp. 523-530, 2017.
[4] H. Kagermann, "Change Through Digitalization - value Creation in the Age of Industry 4.0," in
Management of Permanent Change, Berlin, Springer Fachmedien Wiesbaden, 2015, pp. 23-45.
[5] Porsche Consulting, "Digtial Machinery Decoded - A practical guide for machinery companies to navigate digital transformation and outperform competition§," Porsche Consulting, Stuttgart, 2017.
[6] M. Liukkonen and T.-N. Tsai, "Toward decentralized intelligence in manufacturing: recent trends in automatic identification of things," Springer-Verlag, London, 2016.
[7] A. Shamsuzzoha, M. Ehrs, R. Addo-Tenkorang, D. Nguyen and P. T. Helo, "Performance evaluation of tracking and tracing for logistics operations," Int. J. Shipping and Transport
Logistics, vol. 5, no. 1, pp. 31-54, 2013.
[8] A. Budak, A. Ustundag, M. S. Kilinc and E. Cevikcan, "Digital Traceability Through Production Value Chain," in Industry 4.0: Managing The Digitla Transformation, Istanbul, Springer International Publishing Switzerland, 2018, pp. 251-265.
[9] J. Liua, Q. Wang, J. Wan and J. Xiong, "Towards Real-Time Indoor Localization in Wireless Sensor Networks," in IEEE 12th International Conference on Computer and Information
Technology, Guangzhou, 2012.
[10] J. Werb and C. Lanzi, "Designing a positioning system for finding thind and people indoors,"
IEEE Sperctrum, pp. 71-78, September 1998.
[11] W. Sakpere, M. Adeyeye-Oshin and N. B. Mlitwa, "A state-of-the-art survey of indoor positioning and navigation systems and technologies," South African Computer Journal, pp. 145-197, 2017. [12] J. Figueiras and S. Frattasi, Mobile Positioning and Tracking: From Conventional to Cooperative
Techniques, 1 ed., United Kingdom: John Wiley & Sons Ltd., 2010.
[13] B. Gladysz and K. Santarek, "An Approach to RTLS selection," in 24th International Conference
on Production Research, Warsaw, 2017.
[14] M. Cardullo, "RFIDJournal.com," 21 April 2003. [Online]. Available: http://www.rfidjournal.com/articles/view?392. [Accessed 2018].
[16] Y. Qiao and T. L. Shigang Chen, "RFID as an Infrastructure," Springer, 2012.
[17] T. Watson, "ITAK Information Technolgoy Asset Knowledge," 28 April 2015. [Online]. Available: http://itak.iaitam.org/simple-cost-analysis-for-rfid-options-choice-must-fit-the-organizations-needs-and-budget/.
[18] X. Jia, Q. Feng, T. Fan and Q. Lei, "RFID Technology and Its Applications in Internet of Things (IOT)," IEEE, Yichang, China, 2012.
[19] X. Zhu, S. K. Mukhopadhyay and H. Kurata, "A review of RFID technology and its managerial applications in different industries," Engineering Technology Management, no. 29, pp. 152-167, 2012.
[20] N. Li and B. Becerik-Gerber, "Performance-based evaluation of RFID-based indoor location sensing solutions for the built environment," in Advanced Engineering Informatics, Los Angles, Elsevier, 2011, pp. 535-546.
[21] J. Pancham, R. Millham and S. J. Fong, "Evaluation of Real Time Location System Technologies in the Health Care Sector," IEEE, Durban, South Africa, 2017.
[22] J. Tosi, F. Taffoni, M. Santacatterina, R. Sannino and D. Formica, "Performance Evaluation of Bluetooth Low Energy: A Systematic Review," MDPI, Rome, Italy, 2017.
[23] Bluetooth SIG, "Bluetooth," 2018. [Online]. Available: https://www.bluetooth.com. [Accessed 18 April 2018].
[24] C. Zhou, J. Yuan, H. Liu and J. Qiu, "Bluetooth Indoor Positioning Based on RSSI and Kalman Filter," Springer, New York, 2017.
[25] S. M. Darroudi and C. Gomez, "Bluetooth Low Energy Mesh Networks: A Survey," MDPI, Spain, 2017.
[26] P. Gomathinayagam and S. Jayanthy, "Implementation of Mesh Network Using Bluetooth Low Energy Devices," in Intelligent and Efficient Electrical Systems, Singapure, Springer, 2018, pp. 205-213.
[27] X. (. Shen, M. Guizani, R. C. Qiu and T. Le-Ngoc, Ultra-wideband Wireless Communications and Networks, 1. ed., West Sussex, England: John Wiley & Sons Ltd, 2006.
[28] N. C.-K. Stevan Preradovic, Multiresonator-Based Chipless RFID, 1 ed., Victoria Australia: Spinger, 2012.
[29] M. Raza, N. Aslam, H. Le-Minh, S. Hussain, Y. Cao and N. M. Khan, "A Critical Analysis of Research Potential, Challenges, and Future Directives in Industrial Wireless Sensor Networks," IEEE, Newcastle, UK, 2017.
[30] S. Huanga, Y. Guo, S. Zha, F. Wang and W. Fang, "A real-time location system based on RFID and UWB for digital manufacturing worksho," Elsevier, Nanjing, China, 2017.
[32] S. K. Rao and R. Prasad, "Impact of 5G Technologies on Industry 4.0," Springer, Bangalore, India, 2018.
[33] F. Camacho, C. Cárdenas and D. Muñoz, "Emerging technologies and research challenges for intelligent transportation systems: 5G, HetNets, and SDN," Springer, Mexico City, Mexico, 2017. [34] H. Liu, H. Darabi, P. Banerjee and J. Liu, "Survey of Wireless Indoor Positioning Techniques and Systems," Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews), pp. 1067-1080, 6 November 2007.
[35] A. W. K. V. Chan, D'Ambrogio, G. Zacharewicz, N. Mustafee, G. Wainer and E. Page, "The value of 5G connectivityfor maintenance in manufacturing industry," IEEE, Gothenburg, Sweden, 2017. [36] Federal Aviation Administratio (FAA) U.S., "Global Positioning System Wide Area Augmentation System (WAAS) Performance Standard," 2008. [Online]. Available: https://www.gps.gov/technical/ps/2008-WAAS-performance-standard.pdf. [Accessed 02 may 2018].
[37] Mauro Salvemini, "Global Positioning System," in International Encyclopedia of the Social &
Behavioral Sciences, Rome, Italy, Elsevier, 2015, pp. 174-177.
[38] J. Shen, C. Jin and D. Liu, "A Survey on the Research of Indoor RFID Positioning System," in
Cloud Computing and Security, Nanjing, China, Spinger, 2016, pp. 264-274.
[39] G. Kula, T. Özyerb and B. Tavlib, "IEEE 802.11 WLAN Based Real Time Indoor Positioning: Literature Survey and Experimental Investigations," Elsvier, Ankara Turkey, 2015.
[40] B. Ray, "Airfinder," 2018. [Online]. Available: https://www.airfinder.com/blog/rtls-technologies/real-time-location-systems-overview-technologies. [Accessed 2 May 2018].
[41] D. Kirova, R. Passeronea and A. Ozhiganovb, “A METHODOLOGY FOR DESIGN SPACE EXPLORATION OF REAL-TIME,” Trento, Italy, 2015.
[42] M. Hermann, T. Pantek and B. Otto, "Design Principles for Industrie 4.0 Scenarios: A Literature Review," technische Universität Dortmund, St. Gallen, 2015.
