[In der Blog-Übersicht wird hier ein Weiterlesen-Link angezeigt]

The aforementioned issues are not new and has already been identified during the Big Data era as the Five Vs: Variety, Velocity, Volume, Veracity and Value representing the key dimensions that define and characterize (big) data systems. And these issues are the reason for a low data analysis rate such that only 5% - 20% of the collected test and measurement data is typically analyzed and “less than 0.5% of all data is ever analyzed and used”. 

Addressing the Problems

When looking at the Process of Data Analysis, one can see, that before the actual data analysis and therefore machine learning and AI starts those data problems have to be addressed:

While the picture above (image source) outlines the problem to be solved. There are still many options in the actual how-to – but there’s also an industry standard covering all building blocks for the solution: ASAM ODS. 💙

• Addressing the data format issue
  ASAM ODS introduce the ASAM ODS (External) Data Plugins - lightweight microservices built on the Google gRPC Protokoll, offering a streamlined and efficient API to access both metadata and bulk measurement data—directly from the original files.
• Improving data quality
  Data can then be stored in an ASAM ODS server according to the ASAM ODS data model. This data model adds further data semantic by defining base entities and their respective relations thus defining a measurement data ontology which leads to further machine learning advantages.
• Solving the data access issue
  The designated receiver or user of the data needs to be identified. For data scientists Python has become the de-facto lingua franca for their daily work bridging the gap to various analysis and machine learning libraries, but also to visualization tools such as Microsoft Power BI. Here the open source ASAM ODSBox - a lightweight Python wrapper on top of the ASAM ODS HTTP-API – is helping the data scientist using analysis and machine learning tools such as TensorFlow or scikit-learn, but also Power BI is capable accessing data this way. Checkout the multiple Jupyter notebooks examples in the Data Management Learning Path curriculum to explore the capabilities of ASAM ODSBox.
• Integrating with modern AI ecosystems
  The ODS AIConnectMCP Service provides a unified way to access and share ASAM ODS data for agentic AI workflows. Instead of relying on intermediate formats, ODS AIConnect enables direct, high-performance integration with AI-driven systems. This allows autonomous agents to query, process, and analyze large-scale testing data seamlessly—supporting adaptive decision-making, real-time insights, and intelligent automation across distributed environments.

Conclusion

The combination of ASAM ODS DataPlugins, with the ASAM ODS Data Model including the ASAM ODSBox and the ASAM ODS AIConnector provide the needed functionalities and capabilities making test- and measurement data a first-class ML citizen ready to use in agentic AI workflows.

Outlook

Even though we’re now a huge step forward solving the test and measurement data issues to enable machine learning, the next challenge is already here:  How to include that data in Agentic AI Assistants for test data analysis?

​

💡 Read on in our partner’s Renumics blog how agentic AI system can help to democratize access to this data and to build automated workflows that capture the individual knowledge of engineers.

👉 Contact us to discuss your data strategy.