Digital services become self-learning

Whereas in the 1980s systems operated with fixed logic and delivered the same results for decades, today we expect significantly more from digital services: they should be flexible, adaptive and customer-orientated.

Conventional automation followed rigid processes. Once set up, machines repeated their tasks deterministically – even in cases in which they reproduced errors millions of times over. Modern services, on the other hand, must be capable of continuous learning and adapt to changing requirements based on artificial intelligence (AI) and digital twins.

Within the HARTING Technology Group, the Centre of Excellence Digital Transformation is responsible, among other things, for developing digital services that enable this new type of automation. These services include AI‑supported assistance systems, the Han® configurator, automated engineering‑processes, digital knowledge services and the Group-wide use of digital twins. The aim is to bridge the gap between conventional machine logic and data-driven, intelligent solutions.

At the core is the digital twin, consistently implemented at HARTING based on the Asset Administration Shell (AAS). It provides a comprehensive virtual representation of products, processes and applications and interlinks data from development, production and application. This creates a data-based foundation for automating decisions, deriving variants and intelligently controlling services. The days of pure I/O‑signals are past and gone – modern automation calls for complete, structured data models.

Digitalisation is shifting the focus from merely raising efficiency to actively supporting users. Engineers are not being replaced, but their workload is being reduced in a targeted manner: tasks such as the manual transfer of technical attributes or the remodelling of 3D‑data can be increasingly automated. The less time is tied up in routine activities, the more time will be available for creative and value-adding development work.

In spite of tremendous potential, the widespread implementation of adaptive services is still in its infancy. HARTING is developing the first market-ready applications – such as AI‑supported configuration aids, automated data packages or Digital‑Twin‑based engineering processes. In future, fully integrated, self-learning service chains will emerge. The industry is now in an exciting transition phase in which companies like HARTING are establishing new solutions, step by step.

The transition towards adaptive, digital services is irreversible. AI‑models, time series analyses and digital twins will fundamentally change industrial automation. In future, services will not only automate tasks, but also recognise customer needs at an early juncture and suggest suitable solutions. Performance gains will trend upwards with every interaction – systems will become more context-sensitive, more precise and more flexible.

Data quality remains the greatest challenge. Only when the available data is complete, structured and consistent will digital systems be able to make well-founded decisions independently. But the trend is clear: the industrial automation of the future is digital, adaptive and consistently customer-orientated.