The mixing of granular materials is a critical element in various industries such as mining, agriculture, chemical and pharmaceutical industry. The discrete element method (DEM) [1] is a widely used modelling technique to simulate granular flows and to understand and optimize the processes. DEM treats the granular material as discrete particles and tracks the motion and collisions of these individual particles. However, the large number of particles present in industrial scales leads to increased computational demands, making it less feasible for industrial scale applications. Coarse graining (CG) [2] offers a solution by replacing a collection of small primary particles with larger virtual particles, substantially reducing the number of simulated particles and computational load. However, applying CG to mixing and segregation processes presents unique challenges, as it must preserve key features of the original fully resolved particle system i.e. accurately capturing overall flow and mixing behaviour. In this work, we apply the CG-DEM approach to simulate the current Nauta mixer design, using different CG ratios to balance computational efficiency and accuracy. We examine the limitations of the existing design and explore potential improvements for enhanced mixing efficiency. We investigate the mixing performance by evaluating various mixing indices, as collected in our previous work [3]. Furthermore, to the authors' knowledge, we introduce a novel local mixing index to assess mixing performance at specific regions of the mixer. We also conduct calibration experiments to determine material parameters and validate our simulation results against experimental data. Our findings demonstrate that CG-DEM can effectively simulate the industrial Nauta mixer and provide valuable insights into the mixing process, at much enhanced simulation speed. The validation results show good agreement between simulations and experiments, confirming the reliability of our approach. Furthermore, the local mixing index offers a valuable tool for optimising future mixer designs by identifying inefficiencies.