Case Study: Krones achieves higher Robobox performance and predictive maintenance with MathWorks (Simulink & Simscape Multibody)

Krones Develops Package-Handling Robot Digital Twin

Krones faced the challenge of redesigning its Robobox T‑GS package‑handling system to increase output and flexibility while reducing maintenance for a highly dynamic tripod robot that slides packs across moving belts. To model complex effects such as friction, contact forces, motor inertia, and environmental interactions, Krones worked with MathWorks tools—primarily Simulink and Simscape Multibody (with MATLAB and related toolboxes)—to create a digital twin for design and testing.

Using MathWorks’ Simulink and Simscape Multibody (along with Simulink Design Optimization and other toolboxes), Krones imported CAD STEP data to build a validated digital twin, fitted model parameters to experimental data, injected fault conditions, fed results into FEM analyses, and trained a machine‑learning classifier for predictive maintenance. The MathWorks‑based solution raised robot throughput (the Robobox T‑GS can now group up to 500 layers of six‑packs per hour), shortened product development and testing time, and enables online fault detection and predictive maintenance to diagnose issues before failures occur.

Krones

Benedikt Böttcher

Research and Development Engineer