Internet of Things (IoT) in the supply chain – a jolt of creativity required?


A couple of months ago, I went to an interesting event dedicated to creating a digitalized factory of the future. Given that it took place in Japan, I imagined quite an animated discussion around making another leap in the manufacturing of things. What I got, was quite interesting, but not yet the IoT breakthroughs that I expected.

From the presentations, it was clear that the investment in IoT hardware, from raw material tags, inventory tags, to equipment sensors – is increasing. An automotive brake producer showed how sensors enable better prediction of product wear and tear, simplifying optimization of maintenance. A tire manufacturer supplying to the mining industry provided a similar example of sensors reading and reporting on tire thread status. In another customer example from China, a steel manufacturer –cameras measuring the quality of steel slabs and coiled steel were used to identify non-conformance with the sales order specifications still on the milling line, instead of leaving the discovery of non-compliance problems only at the finished goods lots. Besides those cases, there were of course the ubiquitous examples of autonomous transport vehicles.

Clearly, we have the ways of scanning the tags and interrogating the sensors, but we still require another layer of overseeing control application which then links to the legacy ERP systems. But what I was very keen to see is the concept implementation of an autonomous control tower that goes beyond collating data to inform human decisions.

When devices talk, listen – and act

IoT is supposed to be about devices talking to, and informing each other of what is happening, so that each device in the chain can assess its status and inform other devices of the need to do something, should that self-assessment come back negative.

Imagine a tire wearing too fast telling the vehicle to drive slower or take lesser loads until the replacement arrives, which then in turn warns the downstream machinery to adjust the processing pace to slower/reduced material flow. In essence, the digital value chain should employ all those devices to enable them to be aware of each other, understand their significance within the chain, and allow each device to tell other devices what to do through the data they all stream.

However, at this moment, we seem to be held back by the lack of meaningful chain-wide device integration that does more than simply bring data together. Until today, analysts and many industry pundits talk about extracting device data into some analytical software for the humans to derive some intelligent insights. This frames the IoT discussion as a “big data” problem, detracting from the fact that it should be a “big sophisticated real-time collaborative integration” problem.

On top of that hardware-driven thinking, we still need to think of innovation in the process. I have not seen many examples of creativity in process intersecting with creativity in IoT hardware.

When data has no process support

To illustrate my point, I thought of an example from a year or so ago. To make things more interesting, it is not a factory. I want to show you how IoT could take the guesswork out of preparing the optimal plan and schedule of execution for the entire supply chain in any industry – not just on automated production lines.

Imagine an island-based country with a shared user cement marine terminal. There are four companies using the terminal. Cement companies plan resupply of their inventories in the silos based on demand history. Their chartered resupply vessels (booked 4 – 6 weeks in advance) are scheduled based on port availability to unload their cargo. The cement consumption is driven by the pace of construction on the island.

If a vessel arrives too early and the berth slot is not ready or the silo is too full, the ship anchors and the shipper (a cement company) pays demurrage charges to the ship owner. This charge could be anywhere between $20,000 – 40,000 per day. The vessels arriving on time berth at the terminal within a pre-booked berthing window. As the voyages get frequently affected by weather, some vessels arrive late and miss the unloading windows. Each delayed ship loses its spot and gets sent to the back of the queue. In that case, the shipper also pays demurrage charge to the vessel owner.

The port’s decision to move the resupply ship to the end of the unloading queue disadvantages one of the four cement companies which was expecting that shipment to fulfill committed delivery orders and replenish low inventory in their silo. In fact, the port’s decision could easily tip the market balance in favor of any of the remaining three cement suppliers. It could incentivize any of those three players to temporarily demand higher prices for their product from the distressed buyer. So, the port can easily skew the cement market on the island. This is not supposed to happen in a free competitive marketplace.

In the described situation, we have plenty of software and devices producing heaps of data, but not an optimal market supply process.

A rudimentary management system owned by the port manages the queue, berth assignment and unloading windows, creating data useful for the port’s billing process. The silos are seen through the prism of a minimum/maximum inventory data in the ERP owned by each cement company. And the construction companies, the final users of the cement, have project schedule and cement demand data inside their construction project management applications.

Overwhelming, but underperforming data

In case of devices, on each vessel, port unloading equipment, cement silos, and cement delivery trucks, a plethora of components are already installed as connected devices, but in a limited fashion. For example, ship devices connect back to the ship computer and ship owners, which monitor and control the vessel operations and can optimize their performance individually for the benefit of the ship. The silos have their own capacity sensors. The delivery trucks are equipped with GPS tracking devices. Clearly, an overload of devices and data, and a dearth of optimization. Data from multiple components and systems is not combined and it predicts only a small percentage of supply chain performance issues that arise in day-to-day operations.

Following the old process and thinking, there is not much more to do. But why not think about what smart real-time optimization can do, if we took advantage of all those IoT devices and all that available data?

How optimization balances demand and supply of services

Think of the optimization software placed at the center of the supply network and consider this. Weather on the island drives the construction process.

  • The optimal demand for cement to be mixed into concrete on construction sites can be planned and adjusted daily using a combination of meteorological data, delivery capacity, and construction project schedule data.
  • The cement trucks devices tell the silo when they are coming and the silo devices prepare the optimal loading queue based on the sites’ rate of concrete production. The demand would be balanced against cement supply, both in the silos and in transit on the ships.
  • Capacity sensors in the silos can inform on the rate of inventory depletion, ask devices managing berth access to adjust slots, and tell the ship devices to slow down or speed up sailing to match the new berthing window.
  • The ship’s AIS devices can report the vessel’s sailing progress to the port’s optimization system. In case of predicted ship arrival delay, the optimization software can allocate berth space based on the need to replenish the silos of each cement company, not the queueing order. In case of berth access delay, the port’s optimization system can advise the ship’s engine controls devices to slow the ship down to conserve the fuel. No more need to send the delayed ship to the end of the queue and affect the cement marketplace on the island. No more reason to deny early ship berthing.

Benefits of such different thinking?

  • Increased efficiency of the port
  • Avoiding demurrage costs to the shippers
  • Prevention of price spikes due to reduced supply though any of the four cement companies
  • Sound, vendor-neutral, and competitive marketplace
  • Reduced pollution of the harbor due to shorter anchor times for the vessels awaiting berth slot
  • Reduced pollution in the city from elimination of truck congestion at the silos and at the construction sites

Business collaboration in the digital universe

As you can see from the example above, IoT can spur new business models that would shift competitive dynamics. The combination of process thinking, IoT, and clever optimization software can transform entire supply chains, not only their individual components. A pioneering port could shift their business model to selling information services rather than just access and physical equipment.

Collaboration between the port and all cement companies involved could offer new services to the construction companies. Together, all companies participating in the discussed value chain have enough market power to specify that their IoT vendors make systems interoperable to support their collaborative mission. It’s all about the mission of creating an intelligent digital supply chain where devices talk to other devices, instead of producing siloed data for siloed analytics tools, that still require humans to talk to other humans, thus slowing down the decision-making process and losing optimal outcomes.

Business collaboration clears the way for IoT to become a differentiating factor in industrial competition. But that will only happen, if senior business leaders representing all companies within any value chain take a systems approach to address the business and organizational opportunities that the expansion of the digital universe creates. Without such broad and innovative thinking, the companies will not capture the full range of benefits promised by the Internet of Things.

If you found this topic interesting, leave a comment below. I look forward to hearing what you think.

This post was previously published on LinkedIn. For more posts like this, follow Kris Kosmala on LinkedIn.