Papers

The Physical Internet (PI) has been introduced recently to solve the Global Logistics Sustainability Grand Challenge (Montreuil, 2011). Montreuil et al. (2012) affirm that introducing a new infrastructure such as the PI generates an intense wave of innovative change in business models. This is notably due to the very high level of collaboration that PI implies. Actually, PI consists in a Global Logistics Web (Montreuil et al., 2012) that interconnects all the logistics services through the encapsulation of the goods in smart modular containers (PI-containers) and the use of open logistics facilities (PI-hubs) (Ballot et al., 2012).

Such Logistics Web goes beyond the development of usual Supply Chain networks known in the literature. But to reach this goal, PI should enhance the strategic role of communications and information technologies all along the Supply Chain (Montreuil et al., 2012). As described by (Sallez et al., 2016), PI-containers and PI-hubs must be active to perform adequately in the Logistics Web. This activeness consists in continuously (i) transmitting data that will be (ii) interpreted in order to (iii) support dynamically decisions. In our research work, we suggest to make a parallel between those PI objectives and the current literature on Internet of Everything. Basically, Internet of Things can support the data transmission issue, while Internet of Knowledge and Internet of Services can support the data interpretation and the decision-support issues respectively.

This idea has been made concrete through the MISE (Mediation Information System Engineering) project. This project is dedicated to provide a support framework for collaborative situation by deploying agile mediation information among partners. Obviously, due to its strong collaborative nature, the Logistics Web is a preferred field of application for MISE, even if other application domains have been also developed (e.g. Crisis Management or Healthcare Management). The general principle of the MISE approach is structured according to three steps:

1. Design of collaboration model: this level concerns the gathering of knowledge about the considered collaborative situation in order to instantiate concepts of the so-called collaborative meta-model (concerning mainly environment of the collaboration, objectives of the collaboration, partners and services of the collaboration).
2. Deduction of collaborative behavior model: the second step deals with the automated deduction of collaborative processes, based on the knowledge collected at the previous level. Schematically, the aim is to select and organize partners’ services according to objectives and environment of the collaboration.
3. Deployment of the appropriate Mediation Information System (MIS): the previously deduced business behavior (processes) is translated in a technical behavior (workflows) in order to be implemented. The goal is mainly to match services with activities and data with information.

Furthermore, these three steps are used in an agile framework, which deal with detection of evolution and adaptation of behavior. Performing agility of MIS is based on event analysis (according to the received event, is the situation in line with what is expected) and on behavior adaptation (by invoking step 1, step 2 or step 3 depending on the nature of the event analysis). On a technical point of view, MISE project is based on a Service Oriented Architecture (SOA) paradigm and MISE tools are deployed as web-services on an Enterprise Service Bus (ESB).

In this paper, we illustrate concretely the way MISE might be used thanks to a Logistics Web application dedicated to the agile drug deliveries to French drugstores. Practically, a delivery area is watched through an EDA platform, in order to gather all events (from sensors, services, people, devices, and in a very near future PI-containers) in order to build and maintain a global picture of that area. According to some unexpected (or expected) negative changes (such as an unexpected emergency order, a lot of GPS data showing that a lot of vehicles are stopped, some abnormal values of temperature sensors, etc.), the MISE platform could start the behavior deduction based on (i) information concerning the situation (risk, facts, etc.) and (ii) information concerning delivery means (resources, lead-times, etc.) both extracted from the global picture. Thanks to the implementation step a MIS may be deployed among the stakeholders (drugstores, transportation provider, wholesalers, manufacturers…). Agility of this MIS could be performed thanks to models based on the global picture. Practically, we develop a specific and realistic drugstores’ delivery scenario that shows how the MISE system could provide tooling environment to support parts of the PI and Logistics Web concepts. 

As a conclusion, the MISE platform provides an environment, which allows Logistics Web users to be “hyperconnected” (through topic and content based subscriptions) in order to detect in real-time unexpected events and adapt their behaviours accordingly. One very interesting aspect of that system is the fact that users are not supposed to know each other or even to select the ones they want to get the events from.

References

• Ballot, E., Gobet, O., & Montreuil, B. (2012). Physical internet enabled open hub network design for distributed networked operations. In Service Orientation in Holonic and Multi-Agent Manufacturing Control (pp. 279-292). Springer Berlin Heidelberg.
• Montreuil, B. (2011). Toward a Physical Internet: meeting the global logistics sustainability grand challenge. Logistics Research, 3(2-3), 71-87.
• Montreuil, B., Rougès, J. F., Cimon, Y., & Poulin, D. (2012). The physical internet and business model innovation. Technology Innovation Management Review, 2(6).
• Sallez, Y., Pan, S., Montreuil, B., Berger, T., & Ballot, E. (2016). On the activeness of intelligent Physical Internet containers. Computers in Industry.