AProSys - AI-supported assistance and forecasting systems for sustainable use in intelligent distribution network technology

Climate and energy policy is rapidly changing the energy supply system in Germany. The nationwide integration of renewable energies and the integration of charging stations for electromobility are causing a high level of dynamism that is currently almost impossible to quantify. A forecast of potential outages that adapts to the dynamic power supply system will be necessary in the future in order to ensure the high demands placed on a resilient distribution network, particularly in terms of security and quality of supply. In an intelligent, network-wide energy management system, the rapid response to efficiency losses is also crucial for the sustainability of the distribution network.

Specific recommendations for action must be transmitted interactively to operators and service personnel in real time in order to support and instruct them with activity-relevant and situation-adapted information directly at the systems. The support prepared by AI and made available in the form of digital media directly at the systems "in the field" also makes it possible to provide employees with didactically individualized skills so that, among other things, the challenges of the shortage of skilled workers due to advancing demographic change can be better overcome. Furthermore, travel activities of experts can be optimized and, if necessary, reduced, which also has a positive impact on the CO2 footprint. Especially in times of crisis, as the coronavirus pandemic has shown, digital process support can be an important component in maintaining security of supply. The key to this is an assistance system based on a digital twin enhanced with cognitive capabilities.

The starting point is the sensory monitoring system for medium-voltage switchgear developed in the FLEMING project, which detects technical problems at component level. As part of the AProSys project, an optimized multifunctional variant of this system is being adapted for service life prediction. The integration of adapted forecasting models into the AI-supported assistance system forms the fundamental component, which in particular accurately predicts events relevant to security of supply in the dynamically changing power supply system for a long-term period. AI algorithms based on this provide operators and technical maintenance personnel with prioritized recommendations for action at the line-up level. The assistance system thus not only indicates potential failures, but is also being further developed so that it can provide support in technically complex issues and impart valuable problem-solving skills in the sense of a cognitive system. Another component of this cognitive assistance system is digital support for planning activities within workforce management and knowledge management. Concepts for redesigning services in the distribution network are being developed and validated so that the assistance systems can enable efficient operation and cost-effective maintenance in companies.