Subscribe to our blog
Posted 11 June 2024 by
Koen Cobbaert
Lead Solution Scientist for Axon Technology

Q&A: Axon’s Koen Cobbaert talks probabilistic simulation in supply chain management

In this enlightening Q&A, Koen Cobbaert, the lead solution scientist at Axon, Bluecrux’s digital supply chain twin technology, takes on the intricate world of probabilistic simulation in supply chain management. Through this discussion, we aim to unravel the complexities, benefits and practical applications of probabilistic simulation, offering valuable insights into how this advanced approach is shaping the future of supply chain operations across various industries. This Q&A will give you greater insight into the role of probabilistic simulation in your supply chain and its management and how it can help with planning and optimizing inventory levels, as well as the challenges of implementation and the future trends in the field.

What is probabilistic simulation in the context of supply chain? 

Koen Cobbaert: Probabilistic simulation in the supply chain is a quantitative modeling approach that considers uncertainty and variability in various supply chain parameters. It involves creating mathematical models that mimic the behavior of the supply chain under different conditions. By incorporating probability distributions for factors like demand, lead times and production yields, businesses can run simulations to assess how supply chain performance may evolve over time. This approach enables decision-makers to make informed choices and evaluate the impact of different policies on key performance metrics like inventory, service and cost. 

Why is probabilistic simulation important in supply chain management? 

Cobbaert: Probabilistic simulation is crucial in supply chain management because it helps organizations address the inherent uncertainties and risks that can disrupt operations. By modeling and quantifying these uncertainties, businesses can better prepare for unexpected events, optimize inventory and capacity levels, reduce costs and improve customer service. It provides a structured way to assess the potential consequences of a set of decisions, making supply chain processes more resilient and adaptable. 

What are the key components of a probabilistic simulation model in supply chain management? 

Cobbaert: A probabilistic simulation model typically consists of several key components. There’s supply chain configuration, which includes all the products, customers, suppliers work centers and warehouses. A lot of these data are managed in ERP-systems, although typically a number of other data sources are used as well. 

Stochastic inputs include variables like demand patterns, supplier lead times, transportation times and production yields. These parameters are typically represented using probability distributions.  

The mathematical supply chain model describes how different elements of the supply chain interact and turn the policy into planning. 

Simulation software is a specific software is used to run simulations, generating multiple scenarios by sampling from the stochastic input distributions. 

And lastly, performance metrics, which are defined based on the specific supply chain goals, such as service levels, inventory levels and supply chain costs. 

How is historical data used in probabilistic simulation for supply chain analysis? 

Cobbaert: Historical data serves as a valuable resource for estimating the probability distributions of key supply chain parameters. For example, past sales data can be used to estimate the distribution of forecast errors. Similarly, historical lead time data can inform the distribution of future lead times from suppliers. These distributions are then used as inputs to the simulation model, allowing it to generate realistic scenarios based on historical patterns. 

What benefits can a business derive from using probabilistic simulation in supply chain planning? 

Cobbaert: Businesses can derive several benefits from using probabilistic simulation in supply chain planning, including: risk mitigation, identifying and preparing for potential disruptions; cost reduction, optimizing inventory levels and machine capacity; improved service levels, ensuring timely product availability; scenario analysis, evaluating the impact of various strategies and decisions; and enhanced resilience, building a more robust and adaptable supply chain. 

How does probabilistic simulation help in managing supply chain disruptions & uncertainties? 

Cobbaert: Probabilistic simulation allows businesses to model different disruption scenarios, such as supplier delays or sudden changes in demand. By running simulations, organizations can assess the potential impact of these disruptions on key performance indicators. This information can guide the development of contingency plans and risk mitigation strategies, helping the supply chain remain responsive even in challenging circumstances. 

What role does Monte Carlo simulation play in probabilistic supply chain modeling? 

Cobbaert: Monte Carlo simulation is a powerful technique used in probabilistic supply chain modeling. It involves repeatedly sampling from probability distributions for input parameters and running simulations to generate a large number of possible scenarios. By aggregating results from these scenarios, businesses can estimate the probabilities of different outcomes and calculate key performance metrics. Monte Carlo simulation is particularly useful for complex supply chain models where analytical solutions are not readily available. 

Can probabilistic simulation be integrated with other supply chain management tools & technologies? 

Cobbaert: Yes, probabilistic simulation can be integrated with various supply chain management tools and technologies. The integration works both ways: probabilistic simulation needs to obtain its master data and stochastic information from ERP systems and other sources. The optimized policy can then be fed back to the ERP and/or APS system, so that these systems can plan and operate the supply chain under optimized conditions. 

Additionally, it can be part of advanced analytics platforms that provide real-time insights into supply chain performance. 

How does probabilistic simulation help in optimizing inventory levels in the supply chain? 

Cobbaert: Probabilistic simulation is a valuable tool for optimizing inventory levels in the supply chain by considering various uncertainties and risk factors. It helps in this optimization process in several ways. 

First, there’s demand uncertainty. In many supply chains, demand is subject to fluctuations and uncertainties. Probabilistic simulation allows companies to model different demand scenarios and distributions. By doing so, they can optimize inventory levels to balance the trade-off between excess inventory (holding costs) and stockouts (potential lost sales and customer dissatisfaction). 

Lead-time variability and supplier performance are also major considerations. The time it takes for goods to move through the supply chain can be uncertain due to factors like supplier reliability, transportation delays, and customs processing times. Simulation models can incorporate lead time variability, enabling companies to assess the impact on inventory requirements and determine appropriate safety stock levels. 

Companies often have specific service level targets, such as a desired on-time delivery rate or fill rate. Probabilistic simulation helps determine the optimal inventory levels needed to achieve these service level objectives while considering demand and lead time uncertainties. 

Supply chain disruptions, such as natural disasters, labor strikes or geopolitical events, can have a significant impact on inventory availability. Simulation can model these disruptions and their potential consequences on supply and demand, helping companies plan for contingencies and assess the need for additional safety stock or alternative sourcing strategies. 

Many industries experience seasonal variations and market trends that affect demand patterns. Probabilistic simulation can model these patterns and assist companies in adjusting inventory levels accordingly to avoid overstocking or understocking during peak and off-peak seasons. 

Inventory management optimization often involves minimizing costs while maintaining desired service levels. Probabilistic simulation factors in various costs, including holding costs (costs associated with storing inventory), ordering costs, and stockout costs, to identify the inventory levels that minimize total costs. 

Probabilistic simulation allows for scenario analysis and “what-if” planning. Companies can evaluate the impact of different inventory policies. 

In summary, probabilistic simulation provides a dynamic and comprehensive approach to inventory optimization. It enables companies to consider a wide range of uncertainties and factors that affect inventory levels, allowing them to make informed decisions to balance cost efficiency with customer service objectives. By using simulation, companies can optimize their inventory management strategies to adapt to changing market conditions and supply chain dynamics. x

What challenges should a company be aware of when implementing probabilistic simulation in their supply chain? 

Cobbaert: Companies should be aware of several challenges when implementing probabilistic simulation: 

Top of mind should be data quality and availability. Probabilistic simulations require accurate and comprehensive data on various aspects of the supply chain, such as demand patterns, lead times and inventory levels. Ensuring the quality and availability of this data can be a significant challenge. 

Building and maintaining probabilistic simulation models can be complex and resource-intensive. Companies need to invest in skilled personnel or specialized software to create and manage these models effectively. 

Supply chain parameters, such as demand, lead times and production rates, often involve uncertainty. Accurately modeling this uncertainty is critical but challenging, as it requires statistical analysis and historical data 

Running probabilistic simulations can be computationally intensive, especially for large supply chains. Adequate computational resources, including hardware and software, are necessary to perform simulations in a reasonable timeframe. 

Integrating the simulation model with existing IT systems and data sources can be challenging. Compatibility issues, data format mismatches and security concerns should be carefully considered upfront. 

How can probabilistic simulation assist in evaluating the trade-offs between cost & service levels in the supply chain? 

Cobbaert: Probabilistic simulation helps evaluate trade-offs by running simulations with different cost and service-level parameters. It allows organizations to assess the impact of cost-saving measures, like reducing safety stock, on service levels and customer satisfaction. Decision-makers can find the optimal balance between cost reduction and service level improvement. 

What industries benefit the most from adopting probabilistic simulation in their supply chain operations? 

Cobbaert: In general, industries with complex, global and dynamic supply chains that involve numerous suppliers, varying demand patterns and potential disruptions can benefit significantly from probabilistic simulation. It allows organizations to model different scenarios, assess risks, optimize processes and make informed decisions, ultimately enhancing supply chain efficiency and resilience. This includes industries like automotive, electronics, pharmaceuticals and aerospace, where factors like global sourcing, lead-time variability and demand uncertainty play a significant role. 

How can a company get started with implementing probabilistic simulation in their supply chain? 

Cobbaert: To get started, a company should identify critical supply chain variables and data sources and then collect and clean historical data for these variables. After selecting appropriate simulation software or engaging with simulation experts, a company can configure a simulation model tailored to their specific supply chain. Also key is building a team with expertise in simulation modeling and supply chain management to oversee the implementation and ongoing maintenance of the model. 

Cobbaert: The future of probabilistic simulation in supply chain management is likely to see advances in the use of artificial intelligence (AI) and machine learning (ML). More advanced AI and machine learning techniques will enhance the accuracy and adaptability of simulations. 

With the increasing availability of real-time data from IoT sensors, RFID tags, and other sources, supply chain simulations will become more dynamic. Real-time data integration will enable companies to respond quickly to changing conditions, such as demand fluctuations and supply disruptions. 

Sustainability modelling, considering environmental factors and sustainability goals in decision-making processes, will increase, reflecting a growing emphasis in the area. 

Predictive analytics will be integrated into probabilistic simulation models to forecast supply chain risks and opportunities. This will allow organizations to proactively address potential issues and seize market advantage. 

The interaction between supply chain professionals and AI-driven simulations will become more seamless. AI will assist users in making decisions by providing actionable insights and recommendations based on simulation results. 

In the long term, quantum computing may revolutionize probabilistic simulation by enabling the rapid analysis of complex scenarios and solving optimization problems that are currently infeasible for classical computers. 

What did we learn? 

In this comprehensive conversation with Coebbart, we delved deep into the value of probabilistic simulation in the context of supply chain, tackling the key components and integration with other tools, highlighting the significant benefits it offers in managing uncertainties and optimizing inventory levels. We explored the challenges of implementation, the wide-ranging industry benefits and steps for getting started as well.  

As we look toward future trends, the role of AI, real-time data integration, sustainability modeling, predictive analytics, human-AI collaboration and the potential of quantum computing in probabilistic simulation for supply chain management stand out as groundbreaking advancements. These insights from Coebbart offer a roadmap for companies seeking to harness the power of probabilistic simulation in their supply chain strategies. 

Want to learn more about probabilistic simulation—or Bluecrux’s digital supply chain twin technology, Axon? Reach out to our experts.