Note #14. Machine-learning a Video: “The Corona Virus and the Impact on the Global Supply Chain”

  • Thomas A. Drohan, Ph.D., Brig Gen USAF ret.
  • Commercial, Leadership, Strategy
  • No Comments

This Note analyzes complex linkages in a YouTube video from the MIT Center for Transportation and Logistics: The Corona Virus and the Impact on the Global Supply Chain.

COVID-19 provides an unfortunate example of a ”live case“ that’s relevant to anyone affected by human health. That’s everyone. Moreover, since the C0V-2 virus is a rational actor whose behavior we are still trying to figure out, how we problem-solve this crisis can be applied to other challenges.

Rather than watching the video for 46 minutes actively listening and taking notes, I spent 15 minutes actively thinking with an Artificial Intelligence (AI) platform. Machine and I analyzed relationships being discussed by experts in the video via “supervised machine-learning.” That means I was in charge of the questions, and the AI responded with answers.

Here’s a detailed fly-through and what the sortie yielded. In the end, I was able to quickly grasp aspects of the problem that I would have missed, and think better about what we should and can do in this situation.

Making Sense of Complexity

The takeoff was simple: cut and paste the YouTube video link into the SCAN program of SavantX, a hyper-navigator that discerns relationships in data.  Then we look at the following:

A Big Mess 

What do we see? A complex network of linkages among nodes. Visually scanning the linkages, I notice a relationship between supply chain and vulnerable. Selecting those two nodes generates a new picture with text evidence that shows up as “Results” in the dropdown menu on the right. The text describes how these two ideas are related.

Supply Chain & Vulnerable

Looking at supply chain and vulnerable (two nodes in the middle, above):

  • The discussion is about what supply chain managers should be doing such as find out what supply chains are critical, and which of those are vulnerable. Small suppliers who cannot withstand reductions in orders. Suppliers have too much inventory so they reduce supply to wholesalers several times what may be expected. Wholesalers reduce supply to manufacturers. 

I cross-check the main display again to see other connections to supply chain and vulnerable and see companies as a node. I select companies to investigate any relevant linkages there.

Supply Chain & Companies

Looking at supply chain and companies (the node above supply chain, above):

I want to find relationships being discussed that a company can actually influence now, as well as for future planning. Focus on what we can do, more than what we can’t do, by making decisions and taking action. 

  • Relevant points being made include, how long will supply chains be disrupted? What does customer demand reduction look like? Find out what customers’ past orders were, to see what demand should look like. This can build the basis for estimating the shortfall that needs to planned into current operations.

Ok, but I don’t want to burn gas reacting all of the time; how can I be proactive? I  notice the node, reacting.

Supply Chain & Reacting

Looking at supply chain and reacting (small node to the right of yossi sheffi, above):

  • Most companies react to crisis and need to think ahead more. There’s a need for long term visibility on suppliers, which the Internet of Things will provide with sensor and mapping to suppliers. So we can know where our stuff is coming from. Day to day, before a crisis occurs, few companies are spending the staff—perhaps just one person is needed—to do the long range planning in case of a crisis, or just changes in the market.

Now I search for more opportunities to influence, and notice communication as a node. Even a pilot should be able to influence that…

Supply Chain & Communication

Looking at supply chain and communication (small node in between companies and customer, above):

  • Globalization enables better communication among suppliers, wholesalers, manufacturers, customers. Therefore network now and increase comms for crises.

Ok, got that point. I note shortages. What can we do about those?

Supply Chain & Shortages

Looking at supply chain and shortages (far right, above): 

  • The discussion again emphasizes the need to think ahead about alternate suppliers, deeper-tier suppliers, and remote work to expand a company’s options.

Ok, what about controlling agencies: governments – federal, state, local?

Supply Chain & Governments

Looking at supply chain and governments (center, above):

  • Governments indemnify large business, make loans to small businesses, invest in suppliers. Not anything on levels of government. Bottom line: It all comes down to small suppliers given the networked dependence on particular suppliers from gasoline to meals to cardboard to services.

Too late in a crisis? Not entirely. This one will be long enough for action now to matter, particularly given the exponential rate of spread. Taking advantage of public-private partnerships is important all the time. Decision making  is a node—what’s there?  

Supply Chain & Decision Making

  • The discussion just emphasizes the need to make decisions before we are in a crisis.

As a last maneuver, I deselect supply chain and all nodes, and search the morass again, looking for targets to influence or plan against.

Less of a Mess

I find manufacturing and labor.

Manufacturing & Labor

Looking at manufacturing and labor (center right, above):

  • The advice is to move manufacturing out of China (in this case) to elsewhere, but don’t become dependent on that low-cost elsewhere either.

The point here is wherever we are highly dependent we can be vulnerable, so diversify or know our suppliers’ situation. That’s good general advice, but what are options? What are competitors doing? Finally, I see competitors as a node.


Looking at competitors (small node to the left of cannot):

  • The consensus is that we are all in the same boat, so we must compete to be better prepared.

The remaining relationships display dangerous terrain with respect to competitors: alternate suppliers connected to China; suppliers cannot supply everything. Returning to base.


We spent much more time understanding the complexities of the supply chain problem, and much less on what to do about it. Such is the reality of good mission prep. As for mission execution, there are a number of critical engagements—discern vulnerabilities, specify shortages, improve communications, expand partnerships, adjust manufacturing, remote labor. Overall, there is an overriding need for proactive strategy.

Here’s a checklist of what I learned: 

  • Find out what supply chains are critical to my business, and which are vulnerable.
  • Which suppliers can withstand reduced orders, and for how long?
  • What is the impact on customer demand, and for how long?
  • How can I expand my communication with potential partners?
  • How can I increase my awareness of critical vulnerabilities?
  • Am I creating relationships with alternate suppliers, and cultivating options with current suppliers?
  • Are my public-private partnerships robust enough to take advantage of government programs without being dependent upon them, too?

For a 15-minute session with machine learning, that was time well spent. Human-led AI is useful for problems that require speed and a deep understanding of complex relationships, vulnerabilities and opportunities.

Author: Thomas A. Drohan, Ph.D., Brig Gen USAF ret.

Leave a Reply