The following article tries to describe integration challenges going along with building digital twins of supply chains focusing on maritime environments.
Supply chain disruptions currently caused by a pandemic make it clear that global supply chains are not yet able to withstand massive fluctuations in demand, or only to a limited extent. As early as 1998, during a training course, I got to know the "Beer Game", which simulates the behavior of a multi-level supply chain, which led to whip lash effects when there were even minor changes in demand. One of the main causes of these effects is a lack of holistic transparency and visibility along the supply chain and delayed / limited synchronization of information across supply chain participants. This leads to false demand assumptions by the participants and thus to far excessive orders and thus large fluctuations.
Today - more than 20 years later - there are a large number of supply chain visibility tool providers and Gartner also assesses supply chain visibility as the emerging technology in the supply chain environment with the highest expectations.[1] New technologies such as AI also enable new forecasting options that enable much earlier and, above all, more intelligent reactions to changing needs.
Nevertheless, we have all witnessed massive supply chain disruptions caused by the current pandemic.
Why is this so? Are the supply chain models not yet “real” enough? Are the algorithms not yet mature? Isn't the underlying data good enough?
Even if research into the root causes will occupy us for longer time, it is obvious that in recent years the complexity and length of supply chains have only been taken into account to a limited extent, and the risks that result from them have been underestimated. It seems that too little forecast data was used in the simulation of the supply chains and therefore bottlenecks were recognized far too late.
There are various simulation and optimization tools available - but most of them only cover parts of the supply chain. The implementation of the knowledge gained from such tools can "only" lead to local optima - a holistic view, as being one of the core capabilities of the emerging field of maritime informatics,[2] and thus a holistic optimal across the supply chain cannot yet be achieved. In other words: the closer you want to get to reality and the optimum, the more comprehensive the digital model of the supply chain has to be - thus not just a digital relative but a digital twin.
Abdulmotaleb El Saddik defines:
“A digital twin is a digital replica of a living or non-living physical entity. By bridging the physical and the virtual world, data is transmitted seamlessly allowing the virtual entity to exist simultaneously with the physical entity. "[3]
What are the core components of such a digital twin? Mikael Lind and his fellow co-authors describes the following model in his article "Digital Twins for the Maritime Sector"[4] of which the above definition is providing us with a more fine-grained conception of a digital twin:
Figure 1: The components of a digital twin
While the mathematical models can be configured after an initial analysis and data lakes can be filled with an initial data collection, data streams and situational parameters must be continuously adapted, whereby the boundaries are blurred. In order to be able to reproduce reality as precisely as possible, not only a one-time configuration of the digital twin is necessary, but also a data connection to various real-time data sources. Typical sources for such data streams are sensor data from IoT devices with / and GPS devices. The big challenge is therefore to be able to connect these data sources inexpensively, efficiently and quickly. On the one hand, there are standardization efforts such as UN / CEFACT Smart Container data model[5] and the DCSA IoT connectivity infrastructure[6] to facilitate these integration tasks. On the other hand, due to the different technologies of the systems to be connected, highly flexible tools are necessary with which these connections can be configured and, above all, operated inexpensively, quickly and efficiently.
In his article, Mikael Lind et al outlines the following 3 subject areas as typical application scenarios for digital twins in the maritime environment, the integration aspects of which I would like to address here:
- Fleet optimization at Fright Forwarders
- Terminal, depot and port optimization
- Optimized Container Flow and Supply Chain End-to-End Optimization
In the area of fleet optimization, the connection of fleet management systems is essential in order to be able to map the available resources - from the resource size and shape to the corresponding technical capabilities such as function and equipment. Situational parameters would be the connection to weather data from which e.g. lower driving speeds can be derived, e.g. due to expected precipitation but also road network relevant information such as bridges, tunnels or roadblocks Operational traffic information must also be taken into account and integrated here for realistic planning, whereby these can be assigned to both situational parameters and data streams. Continuous optimization requires the integration of GPS devices in order to be able to track the continuous movements of the resources.
In the area of terminal, depot and port optimization, the connection to their TOS (Terminal Operating System) is a basic requirement. The integration of travel data systems of ships (AIS) and expected arrival times is just as necessary as the connection of camera systems that can scan the license plates of the trucks entering / leaving the entrance gates and plan the corresponding unloading resources based on their load lists. Of course, an integration with HR systems must not be missing in order to be able to plan the really available loading / unloading resources correctly. Ever larger ships have decisively advanced the need for digitization and automated data exchange in this area[7]. Consequently, all transport hubs need to be regarded as transshipment hubs including (sea)ports, and by that not only a particular mode of transports.[8]
In order to be able to digitally record global container movements, more and more containers are being equipped with IoT-based devices. Most of them record GPS position and therefore integrating the underlying systems can create alerts when leaving corridors of geography as well as time. In addition to the global position of the containers, the integration of these SmartContainer enables insights into the condition of the transported goods (e.g. using temperature monitoring) and therefore enables obsolete transports to be identified at an early stage. Informing stakeholders outside the scope of a particular actor is also an important capability of the emerging smartness (of ships, ports etc.) coming out of digitalization.[9]
Conclusion: In order to be able to successfully put a digital twin into operation and to be able to achieve real added value and to be able to switch from a reactive approach to a proactive control of the supply chain, a large number of integration-relevant tasks have to be solved. In order to be able to cope with this successfully, the selection of the right partner is essential, since technologies from different decades from EDI to API have to meet and be connected with each other, especially in the maritime environment.
About the Author
Markus Erdner is focusing on establishing highly innovative international Software companies in the Central European Region with a strong focus on Logistics and Supply Chain focused enterprises. At Youredi he is establishing the Central European Business Unit.
Sources:
[1] Gartner: 2019 Hype cycle Emerging Technologies of Supply Chain Execution
[3] Saddik, A. El: Digital Twins: The Convergence of Multimedia Technologies
[6] www.dcsa.org
[7] Dr Holger Schütt: Digital Twins to improve efficiency by forecasting upcoming bottle-necks
[8] Haraldson S., Lind M., Breitenbach S., Croston J. C., Karlsson M., Hirt G. (2020), The Port as a set of Socio-Technical Systems: A multi-organisational view, in M. Lind, M. Michaelides, R. Ward, R. T. Watson (Ed.), Maritime Informatics (chapter 4), Springer.
