Production process simulations done right

Blog Production process simulations done right

Simulations are a proven tool that can bring value to any business with complex processes and complicated business logic. Whether you are deciding on new investments, searching for bottlenecks or want to test new processes or equipment, simulations can be a powerful weapon, provided they are done right. At the same time, poorly-made simulations will bring very little value to your business, leaving you with an inflexible tool and hard-to-interpret results. So how do you make a flexible and future-proof simulation that shows you valuable insights in the most efficient way?


What is a good simulation?

Generally speaking, a simulation is a deterministic sequential system of moving and stationary objects that interact with each other based on some factory logic. But a good simulation is also:


1)      Flexible, allowing you to change both the parameters of your equipment/machines and the business logic of your processes. Otherwise you are stuck with the initial configuration, meaning you cannot re-use your simulation.


2)      Easy to understand. The output of your average industrial simulation is a bunch of tables and charts. This works perfectly well for some questions like “what is the total time required to tap a section?” and is almost completely useless for more complicated questions like “where and when are the bottlenecks in my processes?”. It’s a well-known fact that humans in general are quite bad at interpreting numbers, but we really excel at seeing patterns in visual data. Yet, for some reason, we keep using the ineffective approach of dumping pages of data and charts at a person, hoping that he or she will be able to discover the right insights from it.


3)      An empty shell of a digital twin of your factory. On some level a simulation and a digital twin are quite similar: they both show you a state of your factory in a given point in time based on your equipment and processes. The key difference would be that a simulation calculates this state, while a digital twin collects historical and real-time data to describe an actual state. This makes simulations the perfect first step towards a digital twin: you replace parts of the simulation with real-life data from your existing systems until you reach a point when you do not actually simulate anything anymore, but have a fully-fledged digital twin. This allows you to generate value with every iteration: the more real-life data you connect, the more accurate your simulation becomes, potentially bringing you new insights about your business. And since you can not manage what you do not measure, we strongly believe that digital twins are vital to reach the top efficiency.


How do we make it?

The key principle of our approach is that we are not building solutions for a particular problem, but we build flexible systems for businesses. Your factory changes with time, and our solutions are ready to change together with your business, so they are never outdated. Here are three specific examples of how we achieve this:


1)   A physics model for all moving objects. Instead of developing separate components for different cranes, AGVs, etc., we have developed a parametrized physics model, that allows us to accurately simulate any moving objects taking into account various settings such as acceleration, deceleration, top speed, and so on. Effectively, this means that it makes no difference for our system if you have one kind of moving machines or dozens of different ones: we can easily simulate them all.


2)   Modular and flexible architecture. All the logic of the simulation is written in a modular way, with each component responsible for a specific part of the factory. We also do not make any long-term assumptions about your factory: what is true today might change tomorrow, so new logic can be easily added and existing rules can be changed at any time. And, of course, any pieces of the simulation can be easily swapped with real-life data.


3)   Visualization for humans. Apart from providing you with summary tables and charts with the simulation results, we also give you an opportunity to see every second of the simulation visualized on your factory plan. This allows you to discover insights in a much more efficient way: gridlocks, non-optimal schedules and routes and other existing and potential issues are right in front of your eyes. This also gives you an unprecedented ability of being able to answer any questions that you might have in the future, while in the case of the traditional simulation results you are usually stuck with the chosen format of presentation (e.g., a table with the end results might not contain that specific piece of information that you are interested in today).

No matter how complex and unique your processes are, we are confident that our flexible system can simulate it. Contact us today to learn about all the benefits of having a flexible and future-proof tool to simulate your processes.

Related updates

Production process simulations done right

Production process simulations done right

Blog
Simulations are a proven tool that can bring value to any business with complex processes and complicated business logic. Whether...
Are plants really getting smarter?

Are plants really getting smarter?

Blog
The difference between automated and autonomous. As a data science company, specialized in autonomous decision making, we have experienced quite...

Back to all updates

Challenge us

Quantillion builds technology for primary metals industry that let every worker, every machine and every plant use data to continuously and autonomously improve safety, quality and output.

If you are interested in what our Industrial AI platform can mean for your organization, we are here to help.

Please input this code: captcha