From Data to Company Values
- PDF / 641,563 Bytes
- 4 Pages / 595.276 x 790.866 pts Page_size
- 55 Downloads / 207 Views
From Data to Company Values AUTHORS
© Sergey Nivens | Adobe Stock
The test process in vehicle development must ensure high quality at maximum cost efficiency while meeting compliance requirements. However, data from simulation and testing all too often ends up in data silos. With a suitable data analytics tool chain, the different types of data can be linked with added value being generated.
Hendrik Bohlen is Technical Director at Werum Software & Systems AG in Lüneburg (Germany).
TOOL CHAIN INTEGRATES ALL TEST DATA
Dr. Stefan Unterschütz is Head of Program Test Process Management at Werum Software & Systems AG in Lüneburg (Germany).
64
Today’s vehicle development makes high demands on the test process. High quality must be ensured at maximum cost efficiency – while meeting compliance requirements. Valuable data on the vehicle are generated in simulations and tests as part of the test process. Among them are tests on endurance, engineering strength, safety and performance. With the use of smart devices, increasing amounts of data are expected to come directly from the field in the future.
Usually, data are stored at the place of their origin, in so-called data silos. Only rarely, they are available across the whole enterprise. That is why it is not possible to derive the added value from them as it could be done by linking these different types of data. State-of-the-art IT offers solutions for the digital transformation of test processes to tap the data’s maximum potential. This article outlines the requirements imposed on a data ana lytics tool chain and a feasible imple mentation. FIGURE 1 shows a simplified depiction of the tool chain process. It illustrates the process steps
data go through on their way from their source to the interested user or continuative systems. The solution can be set up independently of existing IT systems, and the different data suppliers like testbeds, IIoT devices or applications can be integrated gradually. CHALLENGES IN DATA PROCESSING
A state-of-the-art tool chain for the test process has to deliver many functional and non-functional requirements. Once acquired, the test data need to be preprocessed. The data records need to be cleansed, the data are qualified and enriched with additional information or initial statistical evaluations are necessary. In this context, the data may also undergo verification, that is, they are checked for completeness and formal correctness. It may also be sensible to harmonize and standardize the data as part of the pre-processing as it faci litates their future use. Test data require audit-proof storage for the applicable retention periods. Here, accompanying information, socalled metadata, should be filed next to the actual test data. Short access times are desirable to achieve fast processing. Rapidly increasing amounts of data in
today’s test process require a shift to ward scalability. Traceability, transparency and auditing are important elements in dealing with data. Therefore it must be logged how, on which testbed, and by the
Data Loading...
