The Open Systems Interconnection model (OSI model) is a reference model that characterises and standardises the communication functions of the digital internet with regard to its underlying internal structure and technology. Its goal has always been the interoperability of diverse communication systems with standard communication rules. The model comprises of abstraction layers each serving the one above and being served by the one below with the aim to achieve a smooth flow of data.
Physical Internet pioneers like Messrs. B. Montreuil, E. Ballot and F. Fontane proposed that the architecture of the Physical Internet can very well be based on the OSI model, the proposed inspired OLI (Open Logistics Interconnection) model which with its layered stack of services would ensure interoperability and seamless flow of goods within the PI network with information and processes exchanged and executed amongst the layers.
The work in ICONET project is investigating the relevance of the layered Protocol analogies of the Digital Internet to the PI framework as well as the Digital Networking Technologies and Properties as PI enablers. This investigation spans across the PI Framework and Business models work, the Cloud-based Control Platform and Services Design efforts as well as the PI Living Labs testing activities from the very beginning.
More specifically, the OLI Layers investigated and their key defined services /functions are indicatively listed below:
|The Physical Layer||Operations related to the Physical Internet||Identify availability of physical components by requesting IoT data input by sensors or RFIDs (if digitized) The output will be the list of available physical components as per required parameters. Assign physical components to move, unload, load etc, PI containers|
|The Link Layer||Node to node transfer||Gets input from Physical Layer to identify availability of PI Link and PI mean to execute movement. The output will be the available PI links and PI means to execute the specific order as per requirements. Check consistency of physical operations, expose unexpected event/ faults by input of various sensors and events reports and suggest as output corrective actions in the form of new destination PI nodes Trigger notifications to parties involved (FFs/ LSPs, Insurance agents, Destination nodes, Client) of a damaged or compromised PI container from input of events reports. The output will be notifications to parties.|
|The Network Layer||Interconnectivity, integrity and interoperability of networks||Consider key requirements and target KPIs to discover the best network (output will be set of hubs and links) appropriate for the requested order producing the routing table. I-dentify from the routing layer, the main nodes of the network and their functionalities, warehouse capacities, ports, train stations… Assign PI means to PI containers Consume input from external services (e.g. congestion levels on routes) to calculate as output the best alternative for routing decisions|
|The Routing Layer||Routing of the PI containers from starting point to their destination||Calculate best route /path for PI order Identify best route (optimal ones) based on input of criteria /filters including amongst others cost, times…) via a transportation plan (output will be a sequence of segments/ nodes including timing specifications) Provide transit times, transport means, ETAs and ETDs of routing scenario(s)|
|The Shipping Layer||Reliable shipping of PI containers||Provide Shipping instructions, provide status of shipment in the form of notifications and any deviations from original transport plan (in terms of time, cost etc). Output will be details of orders, proof of delivery, date and time and anything relevant to the consignment detail. Transform received IoT data (API key) to Transport Events Expose PI-Shipment Delays/Incident through communication with IoT Devices or other external systems. Output will be an events report Recalculate and expose as output ETA to next PI Node Request services from Logistics web layer (instructions how to proceed), and from the routing layer (updates on the status of the shipment)|
|The Encapsulation Layer||Stuffing /Unstuffing products to PI containers||Assign orders to PI containers and id’s. The output will be a packing list with containers ID and order reference numbers contained within |
Provide information on specific consignments i.e. traceability and how original shipment was converted into a PI one. Output will be a series of events and related PI means and PI nodes to the consignment collected from IoT devices
Provide details of newly constructed PI units (contents) and status
Receive PI order and optimize loading patterns. Output will be a schedule of loading with specific PI units fit to complete a PI container maximizing space
|The Logistics Web Layer||Interface between the Physical Internet and the users of the logistics services||Obtain product characteristics and time specification of order. Input from the ordering party (client) and output a list with characteristics such as weight, dimensions, storage requirements together with the desired time for delivery Provide quotes for shipping goods, times in an optimal scenario. Output will be an offer with best available service to satisfy requirements of the order Dynamic (node to node) cost calculation and revenue as well as distribution amongst nodes and, if applicable, penalties based on SLAs and relative agreements. Create a smart contract (if applicable) Receive and store Transport events in Blockchain|
The interdependence of above can be shown below as OLI layers are re-aligned:
Findings to date confirm that indeed the Protocol Stack can inspire the architectural components of the PI with many shared functional elements. The admittedly complex and interrelated Logistics services of today can vastly benefit from this schema ensuring uninterrupted flow of activities aiming always at preservation of customer service levels and improved efficiency of the supply chain. Minimization of current industry fragmentation and positive environmental impacts are only a few of the associated benefits targeted by the project’s objectives. This is uncharted territory and Digital Internet still differs in some areas to the Physical Internet and therefore further research will ensure a robust and effective PI network.