Performing Fleet Analysis to Meet BS-VI Emission Standards at Honda
“The parallel and distributed computing capabilities in MATLAB enabled multicore processing use of our clusters, thereby reducing the processing time for big data from days to hours.”
Key Outcomes
- Access and process big data of the order of terabytes
- Quickly categorize and filter raw data for feature extraction
- Shorten data processing time by using distributed computing and parallel processing
Honda wanted to create tests for fuel economy and emission constraints to meet Bharat Stage 6 (BS-VI) standards in India. To understand the fuel economy profile for different geographical areas, including variations based on temperature and terrain, Honda engineers had to work with big data sets and perform feature engineering. This work enabled Honda to generate drive cycles capturing real-world driving scenarios.
Using MATLAB®, engineers at Honda parsed through large volumes of raw and unorganized data, extracted meaningful features, and developed models for fuel economy. The processing time for terabytes of drive data was greatly reduced using parallel computing and cluster computing.
With the help of MathWorks engineers, Honda met the BS-VI timelines by using MATLAB for data processing and to scale up. MathWorks training courses also helped Honda engineers quickly attain advanced MATLAB coding skills.
