_____________________________________
* Corresponding author.
E-mail address: behzad.kordnejad@abe.kth.se
Proceedings of 8th Transport Research Arena TRA 2020, April 27-30, 2020, Helsinki, Finland
Internet of Logistics: A New Opportunity for the Digitalization of Logistics
Claudio Diotallevi a , Thomas van Bunningen a , Martin Aronsson b , Behzad Kordnejad c * Jan Bergstrand d , Mats Åkerfeldt d
a
Ericsson,Torshamnsgatan 21, Stockholm 164 80, Sweden
b
PhD, RISE Research Institutes of Sweden AB, Box 1263, Kista 164 29, Sweden
c
PhD, KTH Royal Institute of Technology, Brinellvägen 32, Stockholm 10044, Sweden
d
Swedish Transport Administration, Borlänge 781 89, Sweden
Abstract
This paper explores the applicability of the Semantic Web technologies to the logistics domain, as an open, flexible and efficient approach to share the digital information. The Semantic Web is a framework of technical standards and information management tools that provides a method to store and publish digital data, securely, easily searched, cross-referenced and processed by computers. The proposed approach is introduced within the framework of the H2020 - Shift2rail initiative and FR8HUB project and defined as “Internet of Logistics”, since it aims at creating a common, distributed and interoperable data exchange infrastructure. It overcomes the limitations of traditional supply chain communication systems based on one2one messages (e.g. EDI), and the complexity associated to ad-hoc IT system implementations that are required to build interfaces between legacy systems, enabling new possibilities for the digitalization of processes and operations in the freight sector.
Keywords: Digitalization; Information Management; Internet of Logistics, Shift2Rail
2 1. Introduction
Today, many transport or logistics chain participants are indeed exchanging data between each other, but mostly only in direct peer-to-peer links. There is little or no flexibility to share data with multiple actors or only through data brokers using proprietary business models and technologies with little or no control over with whom and when data is shared and how it is being used and only to a low degree. This leads to a complex and rigid structure for data exchange and reinforces a fragmented, low-margin transport market.
In recent years the application of automatic detection and telemetry technologies have made tremendous steps in availability of devices and reliability of connectivity, particularly in the freight and logistics sector, where shippers, carriers and freight forwarders all aim to achieve full visibility of cargo position, conditions and expected arrival times (ETA) and deviation from plan.
This paper explores the applicability of the Semantic Web technologies to the transport and logistics domain, as an open, flexible and efficient approach to share digital information. The development of intermodal transport is crucial to ensure sustainable growth of Supply Chains that enable efficient Industrial operations.
The Semantic Web is a framework of technical standards and information management tools that provide a method to store and publish digital data. The information sources can be interlinked and more easily searched, cross- referenced and processed by computers. It contains a set of specifications, technologies and data storage solutions that enable the federation of disparate databases that reside on separate servers but share the same common vocabulary and identification mechanism. Such important features allow multiple parties to exchange information without the need to create a common central database and allowing them to stay in control over with whom they share their data.
2. The Internet of Logistics approach
Tim Berners-Lee, inventor of the World Wide Web and Director of the World Wide Web Consortium (W3C) 1 , originally coined the term “Semantic Web” in 2001 2 , while in 2006, he expanded further the concept when he published a note about the Semantic Web project 3 that described the principles of linked data published on the web. The Semantic Web is an extension of the World Wide Web, a collection of standards, common data formats and exchange protocols that “allow data to be shared and reused across application, enterprise and community boundaries” 4 .
The Semantic Web technology provides a convenient solution for the exchange of data among the various enterprises that interact and collaborate to fulfil a cargo shipment, both in logistics chains with low complexity such as a local point-to-point delivery over road as well as in complex, multimodal trade over long-term maritime, road and rail transport or for example fast international airfreight shipments, when multiple transport and service providers or custom authorities need to exchange documents, instructions, invoices, addresses, bills of lading, proof of delivery notes etc. across terminals, warehouses, vehicles/aircrafts, destinations, consignees (customers/receivers) and shippers (senders).
Logistics chains are by their very nature distributed and collaborative so that logistics data exchange calls for the principles of data distribution and collaborative content creation, that have driven the development of the World Wide Web. As in the WWW, the flexible access to information must abide to the limits of personal privacy and the protection of confidential or critical information. Likewise the transport and logistics service providers, buyers and receivers need to exchange information with business partners, but also to protect access from competitors, as well as to ensure the protection of sensitive data (e.g. data regarding dangerous goods).
The proposed approach is defined as “Internet of Logistics” (IoL), since it has the purpose to create a common, distributed and interoperable data exchange infrastructure. As such IoL overcomes the limitations of traditional
1
W3C, International Standards Organization for the World Wide Web (www.w3.org).
2
Berners-Lee, Tim; James Hendler; Ora Lassila (May 17, 2001). “The Semantic Web”. Scientific American
3
Citation (Berners-Lee, Linked Data, Design Issues, 2006)
4
Citation (W3C, 2011)
3 supply chain communication systems that cover only one2one messages (e.g. EDI for direct data exchange or data exchange through data brokers or hubs), and the complexity associated to ad-hoc IT system implementations that are needed to build interfaces between legacy systems. The adoption of web standards, that are part of the Semantic Web framework, will open new possibilities for the digitalization of business processes and operations in the freight sector.
3. Analysis of the approach
3.1. Digitalization in Freight Rail
End-to-end visibility of supply chains is particularly challenging in the international trade that takes place across multiple modes of transport and interchange terminals and that requires a considerable amount of effort spent in hand-over operations between modes, as well as the retrieval and elaboration of shipping documents.
Within the current wave of digitalization projects, shippers and transport operators deploy detection systems and sensors, that are either physically bound to the shipment and the load carrier (container or pallet or package) that are installed on the transporting vehicle (on-board telematics unit) or are physically bound to locations (scanners and cameras).
As an example, the project Shift2Rail FR8HUB/IVG has studied the status in the rail sector, where a relevant share of logistic data is produced by cameras, 3d laser scanners and radio-frequency identification readers (RFID readers), that capture cargo and vehicle features when it passes nearby detection points. These technologies have proven particularly applicable where the installation of such scanning devices along the rail tracks or above road lanes can generate massive amount of valuable data. The project Shift2Rail FR8HUB/IVG has defined the requirements and benchmarks of the so-called Intelligent Video Gateway (IVG) that is a combination of cameras, laser scanners, RFID readers, sensors and possibly thermal scanners, all clamped on a gantry framework above the rail tracks. The equipment registers a detailed “snapshot” of the characteristics of the train and cargo when it passes by the detection point (see picture 1, below)
Figure 1. Intelligent Video Gate installation
The harmonisation of international coding standards for intermodal load units (ILUs) has made it possible to
identify trucks, trailers, containers, swap bodies and rail wagons by reading the alphanumeric codes and icons
printed on the sides and top of a load unit, or that are stored on small RFID chips attached to vehicles (see picture
2, below).
4 Figure 2. Load Unit Image Acquisition and Code detection
The practice to associate coloured placards, alphanumeric signs and RFID chips to load units and wagons is commonly used because this sort of “tagging” comes at low cost and is necessary for handling hazardous goods.
It does not require expensive electronic devices to be installed, powered and maintained on vehicles and containers.
The historical data collected from these kinds of systems in the rail and other sectors can be used to improve forecasting and operations scheduling, based on optimization algorithms. Conversely, real time logistics data can be leveraged to increase flexibility to handle exceptions in a timely manner and to minimize the impact of disruption, such as unplanned delays or handling errors. Increasingly B2B customers count on logistic service providers to ensure predictability of their supply chains, in order to reduce their buffer inventories and allow for a smooth production process. Thus this data, collected about the ILU is a key for improved efficiency and decreased lead times along the transportation chain.
Technical complexity and market conditions discourage the adoption of a traditional IT implementation that would be needed to consolidate logistic data into a central IT system. The freight sector has become increasingly diverse, and it includes many different actors such as shippers, rail network managers, truck operating companies, freight forwarders, terminal operators, customers, customs at country boarders, etc. Various operators share the transport infrastructures and physical resources, but they rarely exchange digital information. A main bottleneck currently to achieve possible business improvements is the lack of end-to-end visibility within logistics chains due to a lack of exchange of data between business partners. Relevant data is increasingly available but it is still challenging for partners to timely share data with each other and consequently all parties are missing opportunities to optimize their businesses.
3.2. How to publish and share Logistics Data with the Internet of Logistics
Unique and shared Identities being web-addresses:
The incredible growth of the internet and WWW has been based on a completely distributed approach, where a distributed network of web pages and documents can refer to one another with global links such as Uniform Resource Locators (URLs, e.g. the blue underlined text that everyone on the internet clicks on to follow the link).
User inquiries are performed by navigating through a vast web of concepts that follows the links across several distributed databases. The global identification mechanism ensures that two pieces of information that reside in separate servers can refer to the same entity. For this the standards of the Semantic Web have defined a global reference that is called Uniform Resource Identifier (URI), and that is somewhat similar to a web address. In the Semantic Web these URIs are used as an “identifier” and “address” for the location where information is stored.
IoL implements the Semantic Web principles and organizes data around digital representations (or digital twins)
of (logistics) objects such as packages, handling units, specific documents and for example locations. These objects
are identifiable with and accessible through a unique identity in the form of a URL (i.e. a web address) that can be
(re)used as unique identifier (URI) by all business partners across supply chains to refer to the specific object.
5 Using URL’s as unique identifiers for digital twins of (logistics) objects allows for truly globally unique identifiers without the need for a central repository or central authority issuing (and charging for) identifiers.
Publish and subscribe to share data between partners and objects
Business partners share data between themselves based on what parties are willing to share with each other. This is done through a publish and subscribe mechanism where data can be shared between partners based on generic rules or specifically for a certain object or group of objects. In this way a consignee of an e-commerce package can for example receive information, such as the Estimated Time of Arrival (ETA) related to his package only, while the driver can access data about all packages on her truck.
In addition, URL’s as identifiers enable any business partner across supply chains to use the URL and “browse”
to the location where relevant data is available and ask for permission to retrieve or add data. This allows for a high level of flexibility with regards with whom data is shared and when. In most logistic chains business partners do not know or do not know up front all other partners that handle their cargo. Using the IoL concept, a ground handler at an airport could for example contact the owner of a damaged pallet and send a photo of the damage by scanning the identifier on the package, getting access to the related data and sending information to the pallet owner without any previous knowhow of, or relation with, the pallet owner. Similarly, this same ground handler could, with a camera, scan all identities that for example are being custom cleared and send a message directly to their owners and other identified parties related to the objects.
Common vocabulary
To overcome the challenge that different data sources and data owners structure and define data in different ways the IoL standard adopts the Semantic Web standard to establish a common vocabulary. The fundamental concept is to structure the information available on the web as a graph of data items, such that cargo, events, documents and pictures etc. are connected by properties that relate such disparate items. A graph database is the most flexible data model as new data can be added and linked to existing records without the need to apply a predetermined rigid schema (as for example with tabular databases with fixed rows and columns). User inquiries are performed by navigating through a vast web of concepts that follows the links across several distributed databases. Large social networks such as Facebook and LinkedIn have being using these graph databases to model relationships between people, friends and colleagues, into so-called Social Graphs. The Web, as a global information space, evolves from linking documents (such as web pages linked with URLs) to linking both documents and data 5 in order to supports the design of innovative applications and relationships between elements of information (Curé, 2015). The 2001 Scientific American article 6 by Berners-Lee, Hendler, and Lassila described this expected evolution of the existing Web to a Semantic Web, to be intended as web of data that can be processed by machines
—that is, one in which the meaning is machine-readable 7 .This opens up for automation of different task, both simpler ones such as arrival checks and validations but also more advanced tasks such as continuously make efficient crane schedules, plan storage of ILUs and (re)booking of ILUs on trains.
As such the Semantic Web enables the access and sharing of information, in ways that are much more efficient and more open than traditional database IT solutions. The Semantic Web supports links between distributed data sources with a structure that is defined and managed by open standards and tools. The World-Wide Consortium (W3C) 8 has defined these tools in the form of standard Semantic Web languages, complete with abstract syntax, model-based semantics, reference implementations and test cases (Hendler, 2011). The standard that the Semantic Web uses to model a distributed web of data, is called the Resource Description Framework (RDF) and it precisely structures the information as a Graph Database, as specified by W3C (W3C, RDF Current Status, 2019).
The Semantic Web provides standard languages and tools to model data in RDF format and to define a common vocabulary, also known as “ontology” that describes the structure of the knowledge. An ontology, or semantic model, takes the role of a glue between federated data sources, so it can be described how they fit together.
Therefore, the structure of RDF databases provides the ability to merge two data sets together and simplifies the
5 From “RDF Database Systems” Olivier Curé and Guillaume Blin – Elsevier Inc. 2015 6 (Berners-Lee, The Semantic Web, 2001)
7 (Berners-Lee, The Semantic Web, 2001)
8 The World Wide Web Consortium (W3C) is an international community where Member organizations, a full-time staff, and the public work together to develop Web
standards. Led by Web inventor and Director Tim Berners-Lee and CEO Jeffrey Jaffe, W3C's mission is to lead the Web to its full potential. Contact W3C for more
information - Copyright © 2018 W3C - https://www.w3.org/Consortium/ (W3C, About W3C, 2018)
6 issue of federating data that are housed in multiple servers, possibly coming from different vendor sources. With this approach, logistic databases will be stored, managed and published at each step of delivery, so that transport operators will establish a web of connected databases where data items (i.e. load units, vehicles, shipping documents, images and 3d scans, train composition, event timestamps etc.) will be structured according to Semantic Web principles.
Other professional communities (e.g. scientists, researchers, media publishers) that need to share and reuse complex knowledge have already started to adopt the Semantic Web standards to establish a common vocabulary, a common “language” that maps the concepts and properties of their domain of interest 9 . Knowledge intensive industries as health care and life sciences (HCLS) provide significant examples of Semantic Web applications, where large numbers of professional users, who work at different organizations in multiple countries, formalize a common vocabulary to share information and to execute projects. Therefore, biology and human sciences research have already proven the validity of the original concept 10 . These Semantic Web applications manage complex knowledge domains such as biology, genomics or drug development cycle. Examples include: The Open Biological and Biomedical Ontology (OBO) Foundry (OBO_Technical_WG, 2018), Gene Ontology (GO) Consortium (The Gene Ontology Consortium, 2018), SNOMED Clinical Terms (source Wikipedia).
Also, in logistics, professional communities have recently started to adopt the Semantic Web standards. The International Air Transport Association, IATA, published their ONE Record Standard to establish a common vocabulary for the air transport industry. The ONE Record standard is based on the IoL concept and is now being tested and implemented by organizations across the globe. IATA’s intention is to work together with organizations from other modes of transport to create a Semantic Web standard based multi-modal data exchange standard.
Usage of the Semantic Web to create common vocabularies allows in the IoL data sources to exchange information instantly (also when previously no data has ever been exchanged), without the need for system integration and without scale limitations to number of partners to exchange data digitally with.
3.3. Application of Semantic Web to Freight Rail
We currently aim at adopting the Semantic Web standards and the Resource Description Framework (RDF) to model information about load units, rail cars and road trucks, and to pilot an RDF database and RDF-based search engine that can merge and publish IVG information to all interested and authorized parties. In order to outline a possible formalization of freight data, in a road/rail context, we model the events and information relative to a load unit, e.g. a container, swap body or semitrailer, that is carried either on road trucks or on rail cars as intermodal terminals perform the transfer of cargo at the end of each segment. The cargo undergoes a series of events at different locations, and detection devices can be deployed at each stage in order to trace the location and condition of the load units (as depicted in fig. 3).
9
Such common vocabularies and languages are called Semantic Models or "Ontologies". RDF data sets can be described with expressive schema languages, that can specify concepts and relationships (that is the “Ontology”); examples of standard ontology languages are RDF Schema (RDFS) or Web Ontology Language (OWL).
10