What role does AI play in your measurement and analysis systems?

You have to differentiate here: Artificial intelligence still plays a subordinate role in our field devices. However, we are observing a significant increase in the importance of AI in the connected evaluation and peripheral devices. The use of AI in control systems and in downstream data systems for process automation, to which our products are connected, is developing particularly dynamically. We are already seeing specific fields of application here, for example in intelligent data analysis, pattern recognition or predictive maintenance strategies.

How does AI help with predictive maintenance?

More and more field devices and components in process automation are now equipped with integrated monitoring functions and can record and transmit additional operating data. This information forms the basis for the use of AI, which makes it possible to recognize patterns and evaluate conditions. On this basis, well-founded statements can be made about maintenance requirements - bothfor the overall system and for individual components - and potentialfaults can be identified at an early stage.

Are there any specific examples from the process industry?

Intelligent applications for condition monitoring are increasingly being used in the process industry. Particularly where measuring systems need to be regularly recalibrated or components cleaned or replaced, AI-supported evaluations enable needs-based maintenance. Instead of rigid maintenance intervals, the optimum time for interventions can be determined in a targeted manner. One example of this is used in the 'Micropilot' with the Radar Accuracy Index, which continuously monitors measurement accuracy and provides early indications of maintenance requirements.

Another practical example is Plugged Impulse Line Detection, a function of Endress+Hauser 's Heartbeat Technology . It is used for differential pressure measurements, for example with the 'Deltabar PMD75B', and detects blockages in impulse lines at an early stage by statistically evaluating the pressure signals - a typical problem in flow measurements with orifices or venturis. Theearly detection of such faults increases the availability of the system, enables condition-based maintenance and ensures safe and efficient operation, especially in demanding areas of application such as chemicals, energy or water management.

How do you combine AI with condition monitoring?

Our current focus is on the reliable recording and structured processing of relevant system data - as the basisfor any condition monitoring. Initial approaches to specifically linking artificial intelligence with condition monitoring are currently being evaluated and are being further developed in close cooperation with our customers.

What data is crucial for your AI models?

For our AI models, the focus is less on classic process measurement values and more on supplementary, so-called secondary measurement values that can also be provided by the measurement technology. These include, for example, temperatures, frequencies, voltages, capacities including time curves or diagnostic information that allow conclusions to be drawn about the health of the device and/or process anomalies. This data provides crucial contextual information and forms an important basis for the development of intelligent, condition-based applications.

How do you deal with data quality?

Ensuring high data quality is a key concern for us - both internally and in our collaboration with partners and customers. That is why we are constantly working on projects to improve internal and external data quality. We rely on established standards such as AAS - Asset Administration Shell -, PA-DIM - Process Automation Device Information Model - or FDI - Field Device Integration -toensurea uniform and structured database.We also actively contribute our experience to specialist committees such as the Industry Alliance and NAMUR in order to promote the exchange of best practices and further develop standards across the industry.

How do you ensure the integrity of the data?

The trustworthiness of data begins in the device. To ensure the integrity of our data, we integrate specific test routines into the firmware that detect manipulations or errors at an early stage. We also rely on proven industrial communication standards and follow the recommendations of relevant bodies such as NAMUR - for example, as part of the PA-DIM data model.We consistently develop newdevices in accordance with the specifications of IEC 62443-4-2, and our development process is certifiedin accordance with IEC 62443-4-1, Maturity Level 3. In this way, we create a reliable basis for secure data throughout the entire life cycle.

How do you help customers get started?

The best way to get started with new technologies is with the right combination of knowledge and practical support. That's why we support our customers not only with information materials such as brochures, white papers and technical guides, but also with targeted services. These include engineering, calibration and maintenance - services that facilitate the transition to productive use and are tailored to the specific requirements on site.

What role does Edge AI play?

In modern automation environments, field devices quickly reach their limits when it comes to storage space or computing power for AI applications. This is exactly where Edge AI comes in: By outsourcing computationally intensive processes to powerful edge devices, complex analyses can be carried out in close proximity to where the data is generated without overtaxing the measuring devices themselves. This approach offers additional flexibility, particularly in the training and test phase of AI models, as updates and adjustments can be made centrally at the edge without interfering with the field device itself.

Where is measurement technology heading with AI?

Artificial intelligence will become increasingly important in measurement technology and gradually establish itself in all components of process automation. It will become a natural functional component that is seamlessly integrated into devices, systems and applications.

At the same time, AI will increasingly support users and operating personnel in their day-to-day work - be it in monitoring, analysis or decision-making in plant operation.

AI also makes an important contribution to achieving goals in production: it helps operators to better implement safety, efficiency and sustainability requirements and at the same time helps tocushionthe effects of the increasing shortage of skilled workers.