Designing your Physical Intranet : Evolving into Digital Supply Chain 2.0

Introduction- The Challenges with Physical Internet

The idea of The Physical Internet was orginally proposed by Prof. Benoit Montreuil in 2011. The link to the original paper is in the Appendix section below. If you are not familiar with the concept, please refer the paper from Prof. Benoit below before continuing. Some additional useful links pertaining to the concept of  the Physical Intranet are also in the appendix.

Unfortunately, the Physical internet is still being discussed as “A New Paradigm” in the world of Supply Chain research. If you search for papaers on this topic, you will see research papaers published as recently as May 2020, calling this theory a new paradigm (see appendix for links). There are annual conferences held to discuss the “Physical Internet of Things” but there is hardly anything to show in terms of real world implementation. The only company, that has made some “internal” effort is Amazon.

Implementing the true form of The Physical Internet means collaboration between companies across Industries to build a shared, common infrastructure. While the fact is that we may still be at least a decade away from it, I think the reason such an innovative idea did not generate the traction it deserved is because we keep thinking about implementing it “AS IS”. I think certain aspects of the idea suggested in Prof. Benoit’s paper can be leveraged in a piecemeal approach and still yield benefits.

We will have to evolve from Digital SCM to Digital SCM 2.0 to build something like this.

Technology and Analytic capabilities required to build a fully functional robust form of the Physical internet do not exist in a mature form currently in my perspective (which shows how ahead of time that paper was). If we are in the era of Digital Supply Chains, we need to evolve into Digital Supply Chains 2.0 in order to be able to implement something like this realistically.

However, companies can start building a foundation in terms of their Digital Supply Chain capabilities, so that they can leverage those capabilities to build the Physical “Intranet” in the future. What those capabilities will be and how you can and should develop them will be covered in a seperate post.

Can Physican “Intranets” be starting points ?

But what if we “customize” the theory to develop a variation of it for internal implemenattion i.e a scenario where each company has its own “Physical Intranet”? And a good aspect of this approach can be that if these intranets are similar to each other, they can eventually at some point connect with each other to create the infrastructure behind the original idea- The Physical Internet.

The most important pre-requisite to venture into this is that your network handles extremely high volume. This makes Logistics Service Providers (LSPs) a very good candidate.

There are other pre-requisties as well, as described in the paper, but in my mind, volume remains a key factor, at least for the Intranet part, since the investment in designing this capability will be very high ($$$). No wonder that the only company that we have heard that is toying with this idea is Amazon, which fits the bill in terms of volume as well as the massive $$$ investment required.

Designing your Physical Intranet

The most fundamental aspect that you need to remember is that this is an internal network.

For each step, I have attached my “Brainstorming diagrams” that provide useful details and also give some examples of how Technology and Analytics will play a key role in designing this primary form of Physical Intranet.

Step 1: Design your standard “Packets”

The first key step is packaging optimization. Now as you may have read, in the original paper, the suggestion is to create packages that “lock together” or snap together. But in a rudimentary form of it, even if we design packaging and standardize it in such a way that the primary packing of most (if not all) of your SKUs are the same….it will still serve the purpose.

Reemeber, the fundamental strength of the mode is that shipments “flow” at package level (much like packets in TCP/IP) so this step kind of lays the foundation of your entire design exercise. If you mus, you can design a secondat packing as well but the best approach will be to keep on form of packaging so that it is the only “product type” that flows in your network.

Then you can start designing other aspects of the network around it…as shown in my brainstorming diagram below. Designing that primary packet or entity that will flow through your network will take into account not only SKU aspects but transportation, sortation, automation aspects as well.

Scan_20200630 (2)

Step 2-A: Start designing the “Information Superhighway”

The packets need to leave your environment and get to a nearest “Router”….in this case, a physical routing hub. Those hubs interconnect with each other to form an actual physical “highway”, much like the “Information Superhighway”.

Remember, another key aspect of this network is that the packet needs to keep moving most of the time.

As packets flow out of manufacturing line, “Intelligent” bots start sorting them to build loads. These packets are directed to pre staged trailers docked at outbound gates. Remember from step 1 that each primary “packet” has an identifier- a Digital tracker which can be in the form of an Internet of Things sensor married with an RFID device.

The conveyor system, specifically optimized around the primary packaing, loads each “facility to hub” straight truck to a high utilization rate. As the packets arrive at the nearest hub, they are sorted as they arrive and go directly into trailers, that have been pre-staged.

Scan_20200630 (3)

Step 2-B: Finalize designing the total end to end “Highway” infrastructure

Now remember that they key aspect here is that continuous flow keeps on happening between hubs i.e there are containers moving continuously. Drivers operate between hubs vs operating by load. What this means is that a driver’s job is to shuttle between say hubs A and B. He visits hub A, drops a container that he hauled from Hub B to A, gets into the yard designated for containers that have hub B bound containers, picks them on FIFO basis.

Note the extremely importat role AI enabled sorting plays so far. In tandem with Packaging optimization, load optimization, schedule optimization, assignment, labor optimization and other optimization tools, the network ensures that loads and containers get filled fast, in a flow format and the wait time of container at a hub is minimal.

Note that with the volume of data that will be generated in these networks, the real time flow directions, load building, container building, load assignment etc. MUST be handled by a Deep Learning algorithm.

Scan_20200630 (4)

Step 3: Design the last mile

In the last mile, if the hub is the destination hub, container “travels” in the yard to a “last mile” processing location. AI enabled bots sort the packets into loads based on the “digital signature/tags” on the packets and loads are built automatically, as shown in the diagram below, leveraging conveyor systems designed around the process.

Scan_20200630 (5)

Conclusion

The concept is still a decade away. And I don’t see an organization transitioning its entire Transportation Network onto suc a platform. Much like the automated DCs, organizations will essentially have to determine what percentage of their SKUs are good candidates for such an initiative and calculate an ROI.

For LSPs, this means a seamless integration with their customers, such that the product flows directly from their plants, as depicted in the diagrams above. But that is only one challenge in developing this.

Technology and Analytics, and advanced AI methodologies in particlualr, like Deep learning (as mentioned above) MUST play a key role to make it work. And to throw a wrench, I don’t think production AI enabled robotics or real life applications of Deep learning have matured to that point as of now.

However, companies can start building a foundation in terms of their Digital Supply Chain capabilities, so that they can leverage those capabilities to build the Physical “Intranet” in the future. What those capabilities will be and how you can and should develop them will be covered in a seperate post.

Appendix

—————————————————————

The Original Paper

The Physical Internet_Original Paper

The basics of Internet routing

We know that the Internet is an open, universal network, composed of countless data packages that move messages from one place to another. When we want to send a message to a recipient,

  • the data we send breaks into this packages,
  • they encapsulate using the IP protocol and are sent to the network.
  • We do not know for which computers the packages pass, the important thing is that they pass through, depending on the time of day, the saturation of the network;
  • all arrive to their destination, where they are detached and the message we sent again is retrieved, so the recipient can interpret it.
  • And to ensure that this works well, there are network protocols, IP addresses, optimization algorithms, etc., which guarantee that no package is lost and that the operation is optimal.

Leveraging the concept to develop Physical Internet

In case of Physical Internet we would do the same but with merchandise sent to a recipient.

  • The merchandise will be transported in modular, standardized and intelligent containers, equipped with intelligent technology that allows each unit to be monitored accurately.
  • These containers have different dimensions in order to optimize the shipments.
  • The products will have to be designed to adapt to this type of containers, to obtain important gains on space optimization and an improvement of merchandise handling efficiency.

Other useful articles

https://www.emerald.com/insight/content/doi/10.1108/IJLM-11-2018-0284/full/html

The Physical Internet: Logistics Reimagined?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s