Preparing the Supply Chain for the Next Disruption

Original article can be found here (source): Artificial Intelligence on Medium

Preparing the Supply Chain for the Next Disruption

By Kartik Pant, Pepe Rodriguez, Olivier Bouffault, and Duane Weeks

COVID-19 has put supply chains for all industries under enormous stress. From the sourcing of raw materials to the distribution of finished products, supply chains are at the frontline of this crisis. But this is not likely to be the last pandemic or global crisis. How supply chains are currently coping — or not — provides some valuable lessons for facing similarly widespread disruptions in the future.

Many firms have taken steps to digitize their supply chains. But many haven’t focused on global supply chain visibility and planning. Faced with the current calamity, companies are frantically diving into supply chain data, creating one-off models to assess risk. As a result, there is very little real understanding of what may happen within a matter of days — which supplier may shut down next, where production may need to stop, or where the next spike in demand will occur.

No wonder, then, that we are where we are today, with supplier delivery and production delays galore, empty shelves of essentials, and pandemonium at retail stores. But it doesn’t have to be this way. With a comprehensive supply chain AI strategy in hand, companies can help avoid similar issues the next time around. Here are some strategies to that end.

Find the Right AI Solutions

First, companies need to take a good look at the digital technologies that they’re currently using to manage their supply chains. Off-the-shelf AI solutions that help with visibility and planning have been around for a while. There are many offerings for specific use cases such as demand management, supply planning, or control tower visibility. But because each offering has its limitations, none is able to provide a comprehensive solution that supports both planning and visibility in both good times and bad.

On one hand, commercial control tower offerings are typically able to provide visibility across the supply chain. But because they don’t provide enough insights and capabilities for day-to-day decision support, they are not the optimal investment in the absence of black swan events. Additionally, these control tower offerings often lack the simulation capability that will allow managers to shift from monitoring what is happening today to predicting what will happen tomorrow.

On the other hand, many supply chain planning tools are great at answering demand-forecasting and inventory-planning questions, but they are not flexible enough to solve for internal and external uncertainty. Their predictive capability is limited to time-series forecasting algorithms, which are quite useful in modeling history to predict variability under business-as-usual cycles. But these same algorithms are insufficient during global disruption because they do not allow for the introduction of internal and external environment variables that are essential for making predictions in crisis situations.

We have supported clients across industries — including pharmaceutical, steel, and consumer goods — in building the initial elements of such a solution. These bespoke elements are essentially visibility and optimization layers that sit on top of standard enterprise resource planning systems to provide transparency and insights for day-to-day supply chain management. But they can be further enhanced to deliver sustained value and better support in times of crisis.

Create Accountability

Getting the right technologies in place is only the first step — the organization piece must also be addressed. All too often, functions such as procurement, manufacturing, logistics, and quality assurance operate in silos, frequently without a cohesive plan to face adversity. While technically the COO’s office provides oversight, there is no clear owner of exchange to exchange (E2E) supply chain visibility and risk resilience. And the potential of advanced analytics to predict and prevent supply chain failure lies largely untapped.

Companies should set up a global team consisting of supply chain strategists, execution drivers, and technologists (including data scientists and data and software engineers). Reporting to the COO, the team should have three primary responsibilities:

· Augment business-as-usual activities across the supply chain. The team should focus on incorporating advanced mixed-modeling techniques that allow nontemporal variables to be introduced. These techniques will help improve the prediction and management of demand and supply during crises.

· Create analytics models. In addition, the team should leverage data and insights from these predictions to develop models that can simulate the sequencing of impact during unforeseen global crisis scenarios.

· Devise guidelines. The team should also put together playbooks that can provide guidance on how to manage the changes to business operations that will be needed in the face of such scenarios.

Invest in a Control Room

Real time (or at least near real time) E2E monitoring of the supply chain is imperative. Now more than ever, companies must set up a control room with live tracking of global suppliers’ health, inventories, cost of goods, production plans, shipments, demand, and external risk hot spots. As the workspace of the global control center team, this room would function as a hub for supply chain optimization during normal operations and as a supply chain command-and-control center in times of crisis.

This capability is analogous to the control room where BCG GAMMA teams have built and deployed AI solutions to optimize tail assignment and crew scheduling for an airline client. These solutions are particularly designed to manage unexpected disruptions, such as delays.

A supply chain control room will require investments in a variety of solutions: track-and-trace technology, including radio frequency identification tags on containers; Internet of Things relays for manufacturing lines; and real-time data processing. It will also be necessary to set up lines of communication and data transmission with suppliers and distributors worldwide. But companies should avoid investing in everything at once. They should first figure out which technologies they already have, what’s missing, the value that the new solutions would provide, and which vendors to procure them from. That will go a long way toward ensuring that these potentially large investments are sound ones.

For example, an industrial and construction equipment company provided its global supply chain control room with visibility into suppliers’ production plans, sales and operations plans, capacity commitments, and inbound flows, as well as into suppliers’ financial, geopolitical, and operations risks. During the 2008 financial meltdown, this visibility proved critical for managing the supply chain and developing contingency plans to mitigate the impact of the crisis. The company did more than survive — it improved its supply chain response time, reduced inventory costs, and increased its market share. This visibility also enabled the company to weather the disruptions caused by the Midwest floods of 2010 and the Japanese tsunami of 2011, when hundreds of suppliers and numerous transportation hubs were out of commission.

Given the critical importance of timely decision making, AI solutions need to cover the E2E supply chain as well as be more granular and near real time than in the past. The challenges posed by today’s interconnected global supply chains call for nothing less.

Use a Digital Twin to Predict and Cope with Disruptions

It’s unreasonable to expect AI to predict when the next global crisis itself will occur. But it can help create a digital twin of the supply chain, a simulation of all the complex systems in the actual supply chain. The digital twin can be used to model what would happen if different disruption scenarios unfold. A well-designed model can help predict demand spikes and supply shocks and recommend what steps are needed to ensure preparedness and resiliency.

In times of crisis, the digital twin can be used to compare the long-term impact of different action plans, making it easier for companies to make good decisions. With the help of advanced forecasting, firms may even be able to detect early signals for what the shape of demand will look like when the crisis stabilizes. Incorporating production planning will make it possible to optimize schedules by ensuring continued supply to minimize delays and maximize financial efficiency. Recommendations coming out of such models may include inventory diversions, customer allocation strategies, and changes in suppliers. Such actions will allow an E2E supply chain to adjust faster as things start getting back to normal.

To leverage the digital twin’s full potential, the control center team will need to be on the lookout to capture internal as well as external data, including how COVID-19 has spread across different parts of the world. It’s also a good idea to extend benefits of such AI-based modeling to suppliers and distributors in the supply chain. Setting up a symbiotic relationship will allow the entire network to benefit from these efforts.

Create a Culture of Disruption Preparedness

Ultimately, for any of this to work, companies will need to establish a culture of disruption preparedness. That means acknowledging the critical role that the control center plays — not only in developing crisis management playbooks and handoffs for all of operations but also in managing change across the supply chain. To shape management policies that are robust, the control center should partner closely with safety and quality functions. Empowering the control center to optimize daily operations will ensure that when a crisis arises, the organization will know how to work with the control team for best results.


The COVID-19 pandemic is not likely to be the last crisis that supply chains will face. But if companies take the opportunity to leverage its lessons, they will have a chance to reduce the damage of the crises that are still to come.

About the Authors

Kartik Pant is a lead data scientist, BCG GAMMA, in the Washington, DC, office of Boston Consulting Group. You may contact him by email at

Pepe Rodriguez is a managing director and partner in the firm’s New Jersey office. He leads our global work in the digital supply chain. You may contact him by email at

Olivier Bouffault is a managing director and partner in BCG’s Paris office. You may contact him by email at

Duane Weeks is a senior advisor in the firm’s Chicago office. You may contact him by email at

Boston Consulting Group partners with leaders in business and society to tackle their most important challenges and capture their greatest opportunities. BCG was the pioneer in business strategy when it was founded in 1963. Today, we help clients with total transformation — inspiring complex change, enabling organizations to grow, building competitive advantage, and driving bottom-line impact.

To succeed, organizations must blend digital and human capabilities. Our diverse, global teams bring deep industry and functional expertise and a range of perspectives to spark change. BCG delivers solutions through leading-edge management consulting along with technology and design, corporate and digital ventures — and business purpose. We work in a uniquely collaborative model across the firm and throughout all levels of the client organization, generating results that allow our clients to thrive.

© Boston Consulting Group, Inc. 2020. All rights reserved. 3/20

For information or permission to reprint, please contact BCG at To find the latest BCG content and register to receive e-alerts on this topic or others, please visit Follow Boston Consulting Group on Facebook and Twitter.