Case Study: Sumitomo Heavy Industries achieves 15% fuel efficiency improvement and 50% reduction in engineering effort with MathWorks Model-Based Design

Sumitomo Heavy Industries Speeds Development of Embedded Model Predictive Control Software for Hydraulic Excavators

Sumitomo Heavy Industries faced a tight deadline to redesign embedded engine control software for hydraulic excavators to meet new emissions standards while compensating for black‑box engines, CAN bus delays, and sudden load fluctuations. To speed development and reduce cross‑team handoffs, the company adopted MathWorks’ Model‑Based Design approach using MATLAB and Simulink (plus toolboxes such as Simulink Design Optimization and Model Predictive Control Toolbox).

Using MathWorks solutions, Sumitomo built a transfer‑function plant model, designed an MPC in Model Predictive Control Toolbox, tuned and verified it in Simulink, and generated production C code with Embedded Coder for a 32‑bit microcontroller. The MathWorks workflow delivered measurable gains: 15% improved fuel efficiency, a 50% reduction in engine speed fluctuations, roughly 50% less engineering effort, and on‑time delivery to meet Tier‑4 Final emissions requirements.

Sumitomo Heavy Industries

Eisuke Matsuzaki

Engineer