The power of AI in predicting the unpredictable
Before the pandemic, many companies prioritized efficiency and streamlined supply chain models, focusing on minimizing inventory and ensuring products were available just in time to meet customer demand. This approach focused on reducing costs and maximizing operational efficiency, with limited buffers for unforeseen disruptions.
Activities focused on reducing waste and increasing efficiency, striving to deliver maximum value to customers and use resources in the most effective way. Operations managers continuously analyzed tasks, processes and personnel to eliminate non-essential activities and ensure seamless communication during every step of the workflow. Digital transformation played a key role in this optimization, centralizing data and increasing visibility, allowing senior leaders to gain greater control over operations.
However, the turmoil caused by the COVID-19 pandemic and more general international instability in recent years have changed the game. These carefully optimized systems, which had undergone iterations of minimal changes to function as efficiently as possible, were suddenly hit by a wave of unpredictable problems.
Lockdowns around the world left certain materials or components in short supply, quarantine times led to transportation delays when crossing borders, and the change in people’s daily lives led to shifts in the demand curve. Furthermore, the rise of remote working forced internal processes to adapt and necessitated new methods of communication and collaboration.
Optimization was no longer the priority as the gains from this approach became negligible in the face of significant potential losses. Instead, the focus shifted to operational resilience. The organizations that came out on top were those with the ability to withstand, adapt to, and recover from disruptive events.
In practice, this means flexible logistics routes that can adapt to geopolitical situations, flexible management of multiple sources and the implementation of tactical buffer plans. The organizations that have successfully implemented these structural changes have managed to avoid customer order cancellations and keep revenues stable.
The disruption caused by the pandemic has taught us an important message: if you spend too much time solving predictable challenges, you won’t have time and resources for unpredictable challenges.
Head of Global Operations, Business Technology and Quality teams, Alcatel-Lucent Enterprise.
A trade-off: efficiency or resilience
Since the pandemic, we have seen some degree of normalization across all sectors. Most organizations are no longer confronted with large-scale disruptions on a daily basis. However, the impact of these extreme events is not forgotten. As operations managers once again look for small profits from streamlined supply chains, the prospect of chaos caused by unpredictable, uncontrollable events looms large in their minds. The question is: how much time should be spent on daily challenges, and how much time should be spent predicting the bigger picture? Limited resources often mean a trade-off between efficiency and resilience – both of which are necessary if an organization is to succeed.
The solution
Automating responses to predictable challenges is the solution to this problem. Technological advancements allow organizations to automate more complex tasks than ever before. Tasks that previously had to be performed manually by employees can now be performed by machines. Not only does this reduce human error, it also allows staff to focus on more complex and fulfilling tasks and leave the mundane to AI or machine learning models.
In the context of the trade-off between efficiency and resilience, automation can be used to solve predictable challenges, maximize business efficiency by reducing waste, and increasing short-term flexibility. AI can handle processes such as sales and operations planning (S&OP), where different business units are coordinated to meet customer demand with the right level of supply. Reporting measures such as demand forecast accuracy (DFA), which measures how well a forecast matches actual demand, can also be automated, reducing the burden on team members.
AI chatbots can be used to communicate with key stakeholders and improve the flow of information internally and externally for an organization. These models can search and analyze vast amounts of data, gathering comprehensive information that can be used in decision-making. For example, a chatbot can instantly communicate inventory availability and plans to partners and customers, as well as the sales team, reducing the need for manual research and back-and-forth email interactions.
Automation assumes regular and uninterrupted processes, which means it is not equipped to deal with irregular events such as pandemics and their consequences. Material shortages, longer transport times and demand instability have an inevitable impact on entire ecosystems. That’s where the human element comes into play.
Technology like AI and machine learning allow employees to spend their time predicting unpredictable challenges and using these predictions to create accurate solutions. The number of possible scenarios means that complex models alongside human judgment and creativity are needed if an organization can devise reactive strategies in advance, allowing it to navigate obstacles on a geopolitical scale. This could mean building relationships with alternative suppliers, building an inventory of key components, or working with customers to propose a staggered delivery schedule in times of unrest.
Come on
Instability is a fact of life, and the modern world will always present us with challenges that seem impossible to predict. The trade-off between efficiency and resilience is not something we will see an end to anytime soon. Organizations must innovate and adapt, leveraging automation to meet the daily challenges of optimization, freeing employees to spend their time and resources predicting the unpredictable. That way, when the next obstacle arises, we’ll be ready.
We list the best OKR software.
This article was produced as part of TechRadarPro’s Expert Insights channel, where we profile the best and brightest minds in today’s technology industry. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing, you can read more here:
