Shekar Natarajan is the founder and CEO of Orchestro.AI.
In March 2021, one of the world’s biggest container ships, the Ever Given, got stuck in the Suez Canal. The ordeal lasted for six days, but the consequences extended far beyond that. A CNBC article explained it well: “The knock-on effects include[d] congestion at ports as well as vessels not being in the right place for their next scheduled journey. Most importantly, it further exacerbate[d] supply chains already reeling from a container shortage amid the Covid-19 buying boom.”
This example is one of many that demonstrates how supply chains are complex, interconnected, interdependent systems. They follow the field-theoretic (geodesic) framework, not the linear approach. The geodesic equation is defined as: “The path or the differential equation of the curve having an external length, i.e., path of extremum distance between two points.” As such, supply chain flows and supply chain AI models are better explained and represented by physics, rather than solely by math.
Why Geodesic AI Models Are A Better Fit Than Linear Ones
At the core, the issue with linear AI models in the supply chain world is that disruptions generally do not move in straight lines—they propagate through curved paths in interconnected systems, creating cascading consequences across different points in the supply chain network.
Consider the issue of a closed port. A linear AI model would use linear effect propagation, approaching the closed port in terms of node-to-node delays—in other words, the closed port would equal a one-day supplier delay, which would equal a one-day delivery delay. But rarely, if ever, is the supply chain world so straightforward. By contrast, a geodesic AI model would rely on causal field propagation, discerning that the same delay would cause upstream shortages, production stoppages and delivery delays at levels of varying intensity throughout the entire network.
Four physics concepts explain why math alone can’t power supply chain AI models: wave propagation, resonance, diffusion and phase transitions.
Like wave propagation, disruptions cause various delays downstream, not just at the point of impact. For instance, a port shutdown for one day in Singapore wouldn’t just cause a one-day delay for ships in that area. It could cause factory delays in India, a wholesaler in New Jersey to be unable to fulfill orders and some shelves of a retailer in Manhattan to sit empty. Approach the port shutdown with linear AI models, and you might call this a node failure with some probabilistic delay. However, geodesic AI models would help you see it for what it truly is—a shockwave that moves through interconnected systems with speed, reflection, absorption and even amplification at various points.
As for resonance, small ripples can have severe consequences. For instance, gate congestion at a port can build up and ultimately cause a shutdown.
Then there’s diffusion. Similar to how terrain shapes how water moves, how a disruption moves through the supply chain ecosystem depends on various factors, such as infrastructure, contracts and relationships between different parties. For instance, if a retail manufacturer’s factory catches on fire, but the owner is friends with another retail factory owner, they could arrange for those goods to be produced at the other factory for a given time period, mitigating the hit to production. However, if a factory fire happens at a facility where the owner doesn’t have such a friendship, the disruption has nowhere to diffuse. That single point of failure causes the impact to flow through limited alternative channels (such as finding new suppliers through formal procurement processes), which amplifies delays and costs.
Finally, phase transitions explain the sudden shifts from stability to chaos in supply chains.
Without Geodesic AI Models, Supply Chains Risk Becoming Increasingly Inefficient
Supply chains today are plagued with various challenges that create inefficiencies and raise costs. Consider a 2023 report by Accenture, which found that “Unforeseen fluctuations in demand at the height of the pandemic, geopolitical unrest driving up wages, material costs and energy prices, climate emergencies, and technology innovations have revealed dangerously low levels of resiliency within engineering, supply, production and operations.” Accenture explained the staggering financial consequences. Specifically, the report noted, “This vulnerability has caused businesses to miss out on a staggering $1.6 trillion in revenue growth opportunity on average each year.”
If supply chain leaders fail to leverage geodesic AI models, they’ll be less prepared to weather the storm whenever disruptions hit. Recovery times will likely be slower, and supply chain systems might collapse under compounding stress. Leaders might miss the signals of disruption propagation altogether. All of these issues can result in companies losing their competitive edge and, in turn, suffering significant financial losses. And in an effort to turn things around for their companies, leaders might find solutions that are optimal for their organizations, but that cause more systemic damage. Or worse, leaders might start behaving irrationally (as is often the case with people facing loss), causing further disruptions in supply chains.
How Supply Chain Leaders Can Start Leveraging Geodesic AI Models Today
By taking steps today to start leveraging geodesic AI models, supply chain leaders can mitigate inefficiencies that are building up in the system.
The first step they should take is to embrace systems thinking, rather than primarily linear thinking, and train their teams to do the same. Supply chains are not a series of linear one-to-one causes and effects. Rather, they’re complex, interconnected webs where a single event can trigger multiple reactions across the network.
Next, supply chain leaders should recognize that a significant portion of the complex web of interactions in supply chains occurs among assets that are rented rather than owned. Additionally, they should understand that variability amplifies through supply chain systems. Linearizing these interactions fails to account for the variability present in supply chain systems. To understand that variability, supply chain leaders need to understand the entities at play, the performance of those entities and the basic variability that each of those entities can introduce.
Third, supply chain leaders should map their networks and build digital replicas of their physical systems—digital twins. By doing so, they can model the physical world and simulate and test how different scenarios would translate to the physical world. It’s also paramount that they invest in real-time data to make this modeling possible.
Finally, supply chain leaders should focus on flow KPIs, namely, lead-time pressure, flow velocity and congestion. If they focus on merely service-level and cost-related KPIs, they won’t get the full picture.
Supply chains are living systems, not spreadsheets that can be perfectly orchestrated, changed and tracked. If leaders treat the supply chains they operate in as physical systems, they’ll be able to navigate anything that comes their way.
Forbes Business Council is the foremost growth and networking organization for business owners and leaders. Do I qualify?
Read the full article here
