Optimizing Reliability Engineering in Manufacturing with Monte Carlo Simulation  

Optimizing Reliability Engineering in Manufacturing with Monte Carlo Simulation  

Sep. 12, 2023
Lumivero
Published: Sep. 12, 2023

As 2023 progresses, manufacturing companies continue to focus on supply chain management issues. In an April 2023 survey conducted by CNBC, only 36% of supply chain managers said they expected inventories to return to normal by year’s end. The New York Federal Reserve’s Global Supply Chain Pressure Index, which collates data from a range of indicators including air freight and shipping costs, began rising again in July after dipping to a historic low in May. With the price of materials and storage still fluctuating, optimizing production processes through reliability engineering can help manufacturers reduce cost pressures elsewhere.

Reliability engineering is a sub-discipline of systems engineering that lets manufacturers fine-tune the dependability of a production plant, process, or finished product, showing how failures or waste in production can affect costs or profits. Predictive analytics solutions that use Monte Carlo simulation can help manufacturers develop reliability engineering models that allow them to effectively manage risk.

Because Monte Carlo simulation can account for a wide range of variable factors, including random chance, it can be tailored to each manufacturer’s circumstances and needs, such as improving supply chain management through optimization of raw material usage or production capacity. This article looks at two examples of manufacturing companies that have made use of Lumivero’s @RISK and DecisionTools Suite to improve their reliability engineering and risk analysis with Monte Carlo simulation.

Met-Mex Peñoles: Testing New Manufacturing Processes Using Six Sigma Design of Experiments and DecisionTools® Suite

Trial runs for new manufacturing processes are a necessary step in research and development — a necessary step that can also be very costly. For Met-Mex Peñoles, cost is even more of a concern. The company is one of the world’s largest silver and zinc refiners and leads Latin America in the refining of gold and lead. Running refinement process trial runs with real precious metals could become very expensive very quickly.

Met-Mex Peñoles uses the Six Sigma Design of Experiments framework (DoE) to develop its trial runs. The DoE framework allows teams to investigate how processes can vary based on:

  • Factors – Different aspects of the process such as heat, pressure, and time in metal refining.
  • Levels – The settings for each factor such as the temperature of a refining process.
  • Response – The output of the experiment, for example the purity or weight of refined metal.

According to Met-Mex Peñoles technology manager Ignatio Quijas, Lumivero’s DecisionTools Suite has played a major role in simulating their Six Sigma DoE trial runs — particularly the Monte Carlo simulation tool @RISK. “Using @RISK to simulate changes in process design allows us to answer some difficult questions without actually running trials,” Quijas told Lumivero.

Using @RISK, his team can input historical data about everything from physical measurements and processing errors to manufacturing equipment tolerances and cost analyses. Plus, the TopRank tool and @RISK’s distribution-fitting feature makes it possible for Quijas to generate graphic representations of simulated trial outcomes that he can use to communicate findings to management teams. The result is less wasted precious metal in trials — and reduced costs for improving their processes.

DecisionTools Demo Request

Szigma IntegRisk® Method: Using Monte Carlo Simulation for Risk Analysis With or Without Historical Data

Met-Mex Peñoles was able to develop its reliability engineering simulations based on historical data. However, what happens when manufacturers don’t have data to use as inputs for their predictive analytics tools? Hungarian risk analysis firm SzigmaSzervíz Ltd. developed a risk-quantification method to deal with just this situation. Their IntegRisk method integrates a scenario analysis process that helps generate input data to inform Monte Carlo simulation models.

Tasked with helping a client understand the potential profit impact of a new plant maintenance process, SzigmaSzervíz asked the project managers to prepare a schedule of activities that would be carried out during the maintenance work. These activities were modeled in Microsoft Project. Assuming no delays, the scheduled maintenance work would take 26 days. Then, managers and team leaders began the scenario analysis. They conducted a series of workshops in which they evaluated each maintenance activity for risk impacts — that is, for any event that could affect the projected duration of the maintenance period — based on their previous collective experience with other maintenance shutdowns.

Using data generated by the company’s scenario analyses, SzigmaSzervíz then ran Monte Carlo simulations with @RISK to generate a lognormal probability distribution. This probability model helped identify both the most likely and the most potentially damaging impact events — for example, a lack of engineering capacity to carry out electrical circuit maintenance activities that could extend maintenance completion time by as much as five days. Then, SzigmaSzervíz developed a tornado graph to show potential delays caused by the most likely impact events and determine what the overall delay to the maintenance timeline could be. The model showed a mean delay of 4.258 days to the original 26-day schedule.

Based on this outcome, SzigmaSzervíz was then able to develop another model that analyzed the potential profit loss that could result from a maintenance shutdown that lasted 30.258 days instead of 26 days. Their findings showed that this delay could cost the manufacturer as much as €5 million. The manufacturer was able to develop risk treatment plans that kept maintenance time close to projected deadlines, ultimately reducing profit loss.

Improve Your Reliability Engineering with Monte Carlo Simulation

Probabilistic analysis tools like DecisionTools Suite can provide manufacturers with a risk analysis process that can improve reliability engineering for production and maintenance processes. Our free example models for different scenarios can help you understand more about how Monte Carlo simulation can help your organization. Download the models today:

Ready to learn more? Watch our free, on-demand webinar Forecasting Asset Renewal and Replacement Using @RISK today!

DecisionTools Demo Request

As 2023 progresses, manufacturing companies continue to focus on supply chain management issues. In an April 2023 survey conducted by CNBC, only 36% of supply chain managers said they expected inventories to return to normal by year’s end. The New York Federal Reserve’s Global Supply Chain Pressure Index, which collates data from a range of indicators including air freight and shipping costs, began rising again in July after dipping to a historic low in May. With the price of materials and storage still fluctuating, optimizing production processes through reliability engineering can help manufacturers reduce cost pressures elsewhere.

Reliability engineering is a sub-discipline of systems engineering that lets manufacturers fine-tune the dependability of a production plant, process, or finished product, showing how failures or waste in production can affect costs or profits. Predictive analytics solutions that use Monte Carlo simulation can help manufacturers develop reliability engineering models that allow them to effectively manage risk.

Because Monte Carlo simulation can account for a wide range of variable factors, including random chance, it can be tailored to each manufacturer’s circumstances and needs, such as improving supply chain management through optimization of raw material usage or production capacity. This article looks at two examples of manufacturing companies that have made use of Lumivero’s @RISK and DecisionTools Suite to improve their reliability engineering and risk analysis with Monte Carlo simulation.

Met-Mex Peñoles: Testing New Manufacturing Processes Using Six Sigma Design of Experiments and DecisionTools® Suite

Trial runs for new manufacturing processes are a necessary step in research and development — a necessary step that can also be very costly. For Met-Mex Peñoles, cost is even more of a concern. The company is one of the world’s largest silver and zinc refiners and leads Latin America in the refining of gold and lead. Running refinement process trial runs with real precious metals could become very expensive very quickly.

Met-Mex Peñoles uses the Six Sigma Design of Experiments framework (DoE) to develop its trial runs. The DoE framework allows teams to investigate how processes can vary based on:

  • Factors – Different aspects of the process such as heat, pressure, and time in metal refining.
  • Levels – The settings for each factor such as the temperature of a refining process.
  • Response – The output of the experiment, for example the purity or weight of refined metal.

According to Met-Mex Peñoles technology manager Ignatio Quijas, Lumivero’s DecisionTools Suite has played a major role in simulating their Six Sigma DoE trial runs — particularly the Monte Carlo simulation tool @RISK. “Using @RISK to simulate changes in process design allows us to answer some difficult questions without actually running trials,” Quijas told Lumivero.

Using @RISK, his team can input historical data about everything from physical measurements and processing errors to manufacturing equipment tolerances and cost analyses. Plus, the TopRank tool and @RISK’s distribution-fitting feature makes it possible for Quijas to generate graphic representations of simulated trial outcomes that he can use to communicate findings to management teams. The result is less wasted precious metal in trials — and reduced costs for improving their processes.

DecisionTools Demo Request

Szigma IntegRisk® Method: Using Monte Carlo Simulation for Risk Analysis With or Without Historical Data

Met-Mex Peñoles was able to develop its reliability engineering simulations based on historical data. However, what happens when manufacturers don’t have data to use as inputs for their predictive analytics tools? Hungarian risk analysis firm SzigmaSzervíz Ltd. developed a risk-quantification method to deal with just this situation. Their IntegRisk method integrates a scenario analysis process that helps generate input data to inform Monte Carlo simulation models.

Tasked with helping a client understand the potential profit impact of a new plant maintenance process, SzigmaSzervíz asked the project managers to prepare a schedule of activities that would be carried out during the maintenance work. These activities were modeled in Microsoft Project. Assuming no delays, the scheduled maintenance work would take 26 days. Then, managers and team leaders began the scenario analysis. They conducted a series of workshops in which they evaluated each maintenance activity for risk impacts — that is, for any event that could affect the projected duration of the maintenance period — based on their previous collective experience with other maintenance shutdowns.

Using data generated by the company’s scenario analyses, SzigmaSzervíz then ran Monte Carlo simulations with @RISK to generate a lognormal probability distribution. This probability model helped identify both the most likely and the most potentially damaging impact events — for example, a lack of engineering capacity to carry out electrical circuit maintenance activities that could extend maintenance completion time by as much as five days. Then, SzigmaSzervíz developed a tornado graph to show potential delays caused by the most likely impact events and determine what the overall delay to the maintenance timeline could be. The model showed a mean delay of 4.258 days to the original 26-day schedule.

Based on this outcome, SzigmaSzervíz was then able to develop another model that analyzed the potential profit loss that could result from a maintenance shutdown that lasted 30.258 days instead of 26 days. Their findings showed that this delay could cost the manufacturer as much as €5 million. The manufacturer was able to develop risk treatment plans that kept maintenance time close to projected deadlines, ultimately reducing profit loss.

Improve Your Reliability Engineering with Monte Carlo Simulation

Probabilistic analysis tools like DecisionTools Suite can provide manufacturers with a risk analysis process that can improve reliability engineering for production and maintenance processes. Our free example models for different scenarios can help you understand more about how Monte Carlo simulation can help your organization. Download the models today:

Ready to learn more? Watch our free, on-demand webinar Forecasting Asset Renewal and Replacement Using @RISK today!

DecisionTools Demo Request

magnifierarrow-right
linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram