Logistical chaos is not a new phenomenon in the supply-chain world, but the scale of upheaval caused by COVID-19 is something that stakeholders have never seen before: 94% of Fortune 1000 companies experienced disruptions from the pandemic, according to Accenture.

Fortunately, businesses today have an array of artificial intelligence technologies such as mathematical optimization at their disposal, to help combat and overcome supply-chain disruptions.

Mathematical optimization has long been established as the go-to technology for supply-chain planning and operations. Since the 1980s, companies across the business world have drawn on a wide array of off-the-shelf and bespoke planning applications, not only to only drive greater efficiency and profitability, but also to manage and mitigate disruptions. Such technologies have been a pivotal tool over the years for fostering supply-chain agility and resilience.

Mathematical optimization facilitates two types of decision-making for planners and other key stakeholders:

• Reactive. They can sense disruptions in real time and respond to them rapidly and effectively by identifying root causes and dynamically reallocating resources, thereby reducing time to recovery.
• Proactive. They can analyze supply-chain risks and anticipate potential disruptions.

Here's how mathematical optimization can drive optimal reactive and proactive decision-making, and address supply-chain disruptions.

A Three-Part Model

Every mathematical optimization application is essentially made up of two elements: a solver (an algorithm-based problem-solving engine) and a model (a representation or digital twin of the real-world operating environment, with all its complexity and challenges).

The model can encapsulate specific elements of the supply chain (such as suppliers, production, logistics and warehouse operations), or it can encompass the entire end-to-end network.

Each model consists of three parts:

• Decision variables. Decisions that are made various points across the supply chain;
• Constraints. Business rules that must be followed;
• Business objectives. Numerous (and often conflicting) business goals, such as minimizing costs and inventory levels, or maximizing resource utilization, on-time delivery performance and customer satisfaction.

When a disruption occurs, mathematical optimization applications, because they’re built on models that understand and embody how an actual supply chain behaves, enables users to achieve:

• Visibility. Instantly identify the sources of the disruption, such as capacity bottlenecks and sudden fluctuations in supply and demand;
• Flexibility. Modify the model by making adjustments and even adding new constraints, decision variables and business objectives to reflect current operating conditions across the supply chain;
• Agility. Dynamically and automatically reoptimize plans and schedules, and determine the best course of action to resolve the disruption as quickly and effectively as possible.

With a mathematical optimization application, companies can maintain real-time visibility and control over the end-to-end network, so that when disruptions hit, they can easily pinpoint the root causes and swiftly take the necessary steps to remedy them and preserve business continuity.

'Continuous Intelligence'

Machine learning, probably the best-known aspect of A.I., relies on historical data. By contrast, mathematical optimization draws on the latest available data to deliver real-time prescriptive analytics, or, as Gartner calls it, “continuous intelligence” across the supply-chain network.

When a severe supply-chain disruption strikes, as it did during the COVID-19 pandemic, companies can’t depend on data from the past to help navigate the unprecedented financial and operational challenges.

Because they utilize the latest available data and models that capture current conditions across an operational network, mathematical optimization applications are capable of automatically generating the best solutions to present-day supply-chain problems, and enabling continuous intelligence and optimal decision-making.

Exploring Risk

An important part of handling supply-chain disruption is assessing risk and planning and preparing for the future. With mathematical optimization’s scenario-analysis capability, companies can:

• Explore various supply, demand, inventory, capacity, macroeconomic, geopolitical and other what-if scenarios, and evaluate their potential effect on the business.
• Uncover hidden risks and gauge risk exposure and time to recover in the event of a disruption such as a natural disaster, or production or transportation breakdown.
• Unlock opportunities to mitigate risk and drive improved supply-chain resilience, by reallocating resources or reconfiguring the supply chain.

Mathematical optimization’s scenario-analysis functionality allows companies to insulate their supply chains against the impact of future disruptions, by enabling proactive, strategic decisions in multiple areas, including capital investment, supplier selection, capacity and inventory planning, and production and warehouse facility location.

During the COVID-19 pandemic, we’ve experienced an unprecedented wave of supply-chain disruption, which has caused significant and lasting chaos in the global economy and posed huge challenges for supply-chain professionals. Mathematical optimization has proved itself a potent weapon for battling such disruptions, while boosting supply-chain efficiency and profitability. This A.I. technology will continue to be an essential tool for supply-chain leaders as they navigate the ever-changing business landscape.

Ed Rothberg is co-founder and chief executive officer of Gurobi.