Managing supply chains entails many types of practices – transportation and logistics planning, material flow planning, manufacturing planning, distribution planning, macro supply chain planning, etc. So-called supply chains are large and complex ‘organisms’ that require the work and involvement of people in many Firms of diverse specialties. It’s not one thing and it’s very complex.

What is common to all these practices, though, are systems. Managing supply, at any level, can’t be done without information systems.

Here’s the problem: the information systems that became common place (if not standard practice) across supply chain players are all architected using approaches to systems engineering that are no longer valid in today’s world of highly volatile supply chains where unpredictable events and unpredictable behaviours are the norm.

In the world of supply chain, a system is always based on the same basic principles:

• You represent the outside world with data structures and parameters
• These data structures and standard parameters are assumed to be an adequate model of the real world, i.e., the data is assumed to be reasonably stable. Take lead time: you indicate the lead time of a supplier for a product and then keep it that way for months or years.
• What supply chains systems do a lot is planning; and all forms of planning are made of algorithms based on the said parameters and data structures. These algorithms are typically linear with logic that assumes a relative predictable and stable order of things in supply.

The New, Unstable ‘Order’ of Things

The “stable order” I’m referring to is what any supply chain Firm – whether manufacturer, distributor, retailer or service provider – has been accustomed to for decades:

• Relationships in a typical supply chain of that industry – types of suppliers and clients, the role that each entity plays in the chain, the types of orders that are placed between players, transportation routes, transportation modes, borders, border crossing protocols, etc., etc.
• Where you procure supply – whoever you are in a supply chain. What you get from India, China, South America, Mexico, Europe, etc.; what you import and what you make locally; when is best to subcontract; etc., etc.
• Who are your Clients; the types of contracts and relationships you have with them; who are your competitors, what are their strengths, how you compete with them.

The “stable order” of things in most supply chains has been under threat for at least a decade and a half, particularly with Digitization and the emergence of economic power in Asia. That, however, has been gradual.

The Pandemic introduced a number of factors that started accelerating the transformation of supply chains. Gradually but surely, supply environments became more and more uncertain to the point where the usual order is no longer the usual order. Nothing is usual any longer.

During the pandemic, uncertainty of supply effectively reconfigured supply chains over and over as providers had to look for new suppliers and partners, lead times were all over the place, shortages in the market became endemic – and this over short bursts where new decisions had to be made quickly and frequently.

The pandemic also left us with a world in chaos because of all the changes that accelerated so drastically in 2020-2021:

• Digitization and Work from Home (WFH)
• Remote work that provided access to human capital across vast distances and political boundaries
• Rapid emergence of new technologies in health care and in materials science
• The Digitization of transportation and the emergence of virtual supply chains
• The emergence of blockchains to the mainstream
• The emergence of crypto – both the good (speed and convenience) and bad (an epidemic of scamming)
• The emergence of distributed micro factories
• The employee ‘rebellion’ against oppressive work environments that led to the Great Resignation and is now causing staff shortages in every supply chain.

Overlaying these general phenomenon, you have macro economic factors that are rattling the whole world, such as:

• The Ukraine war
• Insane volatility in energy costs
• The impact of Climate Change – droughts, floods, forest fires, heat waves – which contribute to paralize large swaths of society in many parts of the world.
• The electrification of transportation – personal, public and commercial
• Expansion of satellite communications

…to name but the main ones

Now: imagine that you are a planner in a supply chain – anywhere in a supply chain – and you need to represent this world with static structures and parameters that don’t change – good luck with that. In addition, the algortitms of supply chain systems – like a standard MRP materials planning algorithm – are too simplistic to deal with everything that affects the flow of supply.

So: all these systems that permeate the industrial world are thoroughly inadequate. They simply can’t cope with this crazy world of ours and, unfortunately, the crazy world is only going to get crazier.

What do we do, then?

Classical information systems can’t cope with today’s volatile world but we’re not going to throw them out. They do things that need to be done so – they’re there, we know what they can do, let’s leave them there and come back to them after we dealt with volatility.

What I recommend is a perspective of supply chain operations with 4 strategies. Through these 4 strategies, we deal with volatility but we also try to keep operations relatively stable, at least for short periods of time. If operations are being rattled and rescheduled and rearranged constantly, costs will go through the roof and companies will become non-viable.

Strategy 1 – People

Strategy 2 – AI

Strategy 3 – Simulators

Strategy 4 – Operational Systems

1 People

Let’s say you’re looking at a company with 10,000 people. They’re not just 10,000 people. They are a network of many people, interacting amongst themselves in a myriad of ways. It’s through these interactions that a company delivers to its customers the products and services it promises.

So, this network is going about its business and volatility strikes. They need to react and fast. The longer it takes to react, the worst the impact of the disturbance. What is the ideal behaviour for a network of people subject to frequent volatility?

Think about Swarms. Swarms of birds, wildebeests, fish, nanomachines – anything. A Swarm is made of living, intelligent animals who function as a collective, i.e., with collective intelligence. When there is a disturbance, the entire swarm moves rapidly, completely in synch. There is no central command telling each animal or person what to do. They just move as a whole.

Human beings are susceptible of and capable of swarm behaviour. We see this in elections when there is a tipping point and a large portion of the electorate swings in one direction. We see it in natural disasters when people step out as a whole and go help and pull themselves together and intervene as a group.

We also see it in organizations. Companies that are structured as Agile models are capable of quick, collective reactions, similar to Swarms, without direction from a central command authority. You can read about it here.

When Agile networks of people extend beyond the boundaries of one company and connect with people in other companies, you may call that “collaboration”.

Thus, the most responsive supply chains, in conditions of high volatility, are those who are made of networks of people in multiple companies functioning in Agile models.

2 AI

I’ve written extensively about the role of AI in volatile supply chains. You can find my thinking in this article.

In essence, AI can do what classical systems cannot. AI is a technology where a computer learns through experience, where that experience is contained in a large data base of structured and unstructured data that we call a Big Data lake.

AI models ‘gobble’ large loads of data about a supply chain environment, including whether, politics, transportation routes, harbours, pandemics, natural disasters, as well as data about suppliers, regulatory entities, etc. All this data enables the AI model to predict demand and supply more accurately than the simplistic linear model of classical systems.

3 Simulators

In the world of traditional systems, there is a category that I will call “Simulators”. They are sometimes know as “APS” or “Advanced Planning Systems”. Some APS’s can simulate, some can’t. The best are made for end to end simulation of supply chain scenarios.

A good simulator can take the predictions of an AI model and translate them into the detailed implications on supply operations – raw materials manufacturing, finished goods manufacturing, distribution, etc.

It’s important to have a really good simulator so that you can more precisely translate AI predictions into logistical implications on work schedules, labour capacity and machine or space capacity.

4 Operational Systems

And… Here we are – with classical operational systems. These systems (the best known in the industrial world in SAP) are quite complex but they support action in the field. They support administrative tasks that need to be done in order for the supply world to function.

Of these 4 strategies, the most important one is the first. If you don’t transform your organization into an Agile network with swarm-like behaviour, everything else is a waste of time and money. Conversely, you can create an Agile network of people without touching any systems.

Agile and AI are perhaps the two strategies that most people are unfamiliar with. They are, however, simple. They may seem scary to some but that’s just because they are unknowns. Once you know them, you will see they are actually quite simple and even intuitive.

In conclusion: managing supply chains in a highly volatile world is possible, using 4 straight forward strategies as a series of layers that feed into each other. You can see 1 to 4 as a ‘funnel’ whereby the volatility experienced by the company is translated rapidly into operational activity. As you go through the funnel, people must take care to change as little as possible so as not to allow disruptions at the periphery to permeate without order into the core of the enterprise.

This Mesh of people networks combined with AI, simulators and operational systems is, in my current thinking, the best way to return to reliable supply without waiting for the world to “settle down”… Because it’s unlikely that it will settle anytime soon.