Brownfield Factory Layout Planning using Realistic Virtual Models

Doctoral thesis, 2021

To stay competitive in an increasingly digitalised and global context, manufacturing companies need to increase productivity and decrease waste. This means their production systems must improve; something they can achieve in a multitude of ways. For example, increasing the level of automation, improving scheduling and improving product and process flows. Often, these production system improvements entail redesigning the system to incorporate these ensuing changes; a unique and temporary endeavour that is often structured as a project. One part of the production system design process is layout planning, in which the positions of operators, workstations, machines and other parts of the system are decided. This planning process can have a major impact on the overall efficiency of operations.

In industrial settings, factory layout planning is often conducted in brownfield settings. In other words, in operational facilities. Since every production system and facility is unique, so is every factory layout planning project. Each such project has different preconditions, existing knowledge, availability and quality of data, lead-times, expectations and driving forces, to name just a few. If factory layout planning were treated as a design problem (more subjective than mathematical in nature), it would be hard to produce a mathematical solution for an optimal layout that would also work in reality. Instead, if a layout is developed and adapted to all real constraints and factors while it is being developed, the result would more likely be installable and work as expected.

The long-term vision of this thesis is of a future in which sustainable manufacturing industry continues playing a vital role in society, because its contribution is more than just economic. A future in which the manufacturing industry is appreciated and engaged with by the local community; in which high performance is connected to the successful adoption and efficient use of digital tools in developing and improving existing brownfield production systems. This thesis aims to ensure that manufacturing industry adopts realistic virtual models in its brownfield factory layout planning processes. It does this by identifying and describing common challenges and how they may be reduced by developing and using realistic virtual models. This leads to improvements in the planning, installation and operational phases of production systems.

The findings of this thesis show that brownfield factory layout planning represents a significant proportion of industrial layout planning. Its challenges lie mainly in the areas of data accuracy and richness. There are difficulties in grasping scale and perspective, communicating ideas and gathering input in the layout planning phase. By applying 3D laser scanning to provide accurate data and virtual reality to provide immersion and scale, realistic virtual models have been created. These reduce or eliminate the challenges stated above and allow more employees to be involved in the layout planning process. This, in turn, results in the identification of flaws in the layout and improvements in the early stages, rather than during or after installation. There is also an overall improvement to brownfield factory change processes, with costs that pale by comparison to the total cost of layout changes.