Digital Twinning is a term which means creating a digital model of a physical system and nurturing it until it is a clone of what happens in the real world. Digital Twins are best when they have been validated by using a good monitoring system.
Bullant Filters are an example of a spreadsheet-based digital twin for a processing line. However, specialising in the process industry means that we are often faced with systems that are so complicated, and include so many interactions, that it becomes impossible to optimise them by using just a spreadsheet. In these circumstances, the only way to verify and improve performance is to model and monitor the system using specialist software.
We use discrete event simulation software for complex factory and warehouse optimisation tasks. This means we can use Digital Twins to confirm performance and productivity benefits of project design proposals before committing to high-cost assets or people.
We can provide a cost effective Digital Twinning strategy using the following;
Factory Dynamic Modelling
Model a factory or production line to identify buffer requirements or points of highest performance leverage before committing to expensive assets or costly improvement resources.
Use data to monitor the performance of different parts of your supply chain. Compare monitored performance with modelled performance to optimise. Overlay machine learning to warn of potential system changes.
Warehouse & Logistics Dynamic Modelling
Warehouses can include very complicated interactions between material handling equipment such as conveyors, robots, and forklifts. In these situations the cost of running a warehouse facility is determined by the way employees (such as pickers or forklift drivers) interact with the equipment.
Often the only way to identify optimum layout and minimise running costs is to model the warehouse system using a computer model.
We discrete event simulation software to model a warehouse system and show what-if scenarios for different equipment and employee configurations.
Sometimes a business must consider options for reconfiguring a supply system and this redesign often involves changing the location of manufacturing facilities or warehouses relative to customers.
Small differences in location may constitute large changes in cost and time and this can have a huge impact on total Cost-To-Serve and customer service. Often, the only way to demonstrate this properly is to construct a simulation model showing truck movements.
We use the discrete event simulation software to simulate truck movements for different warehouse and supply facility location alternatives.
Factory Dynamic Modelling
Sometimes buffers (such as conveyors, tanks and silos) are required to create a particular operating and performance outcome for a factory. In these situations it is important to ensure that buffer sizing and location is as accurate and streamlined as possible to minimise asset costs while achieving the business outcome.
Also existing high speed, high technology production lines consist of a series of workstations with differing equipment capacities, buffer sizes and buffer positions. This can make it difficult to identify the most powerful performance leverage points in a line.
In these situations, we use discrete event simulation software to dynamically model a process so that performance and productivity benefits can be confirmed before committing cash or resources.
Using dynamic factory modelling we can;
- Identify constraints and performance opportunities within complicated flow-line operations.
- Define optimum buffer levels between processes.
- Determine constraints caused by resources such as forklifts or employees.
- Enable design verification prior to capital expenditures.
- Enhance decision making, particularly in relation to buffers.
- Reduce the risk of costly design mistakes by running what-if scenarios.
Monitoring operations across many sites and processes can be difficult. There are 4 issues that make this a challenge:
- the quality of data can vary from one site to another.
- the way that metrics are calculated is at the mercy of different interpretations for each site
- metrics (such as OEE) are difficult to relate to business outcomes
- solutions usually require a large, all-or-nothing, capital outlay
We have developed a method of performance monitoring that overcomes these issues and requires, on average, less than 4 inputs for each product run.
Information from this monitoring system can be used to:
- compare modelled performance with monitored performance and so identify opportunities for improvement
- compare different sites and processes using like-for-like metrics