Delete search term


Quick navigation

Main navigation

Improving the trustworthiness of critical systems with AI: ZHAW and FHNW are part of the EU-HORIZON project AI4REALNET

In the AI4REALNET project, researchers from the ZHAW School of Engineering, the FHNW School of Applied Psychology and the Swiss Federal Railways (SBB) are investigating the interaction between humans and AI-based solutions for critical systems such as electricity, railroads and air traffic control. A central question is: What technological and ethical challenges arise from this human-AI cooperation?

The EU project AI4REALNET is a collaboration between the ZHAW Centre for Artificial Intelligence (CAI), the ZHAW Institute for Data Analysis and Process Design (IDP), the FHNW School of Applied Psychology, the SBB and several international universities and industrial partners. AI4REALNET thus brings together a wealth of expertise from eight countries and was selected along with three other projects from 114 submissions as part of a highly competitive European call for proposals.

ZHAW and FHNW join forces for human-centered and robust AI

Critical infrastructure networks for mobility or electricity are usually operated by humans, but increasingly human expertise is being supplemented by control and monitoring software and various degrees of automation. "As we are dealing with sensitive infrastructures, the stakes are very high. The AI systems must be reliable so that the critical applications are not jeopardized," says Ricardo Chavarriaga from the CAI. For this reason, a ZHAW team consisting of Thilo Stadelmann, Manuel Renold and Julia Usher will implement a powerful method known as reinforcement learning, which adapts to challenges. At the same time, a team consisting of the ZHAW researchers Christoph Heitz, Ricardo Chavarriaga and PhD Student, Wolfgang Stefani – co-supervised by Teresa Scantamburlo from the University of Venice – is working on the question of how the ethical aspect of the interaction between humans and AI can be formalized and addressed in the context of critical infrastructures. The challenges for cooperation between humans and AI arise, for example, from the increasing uncertainty due to weather, age of the systems or demand.

The project team at the FHNW School of Applied Psychology, consisting of Andrina Eisenegger, Samira Hamouche and Toni Wäfler, is focusing on the research objective of "co-learning". The aim is to understand how humans and AI can interact and learn from each other. Toni Wäfler, Professor of Work and Organizational Psychology and project manager at the FHNW, says: "From a psychological perspective, the aim is to analyze how specialists work today. What skills and needs characterize them? From this, we can derive how AI can be profitably integrated into their work processes."

Robust operations for rail, aviation and energy

The main objective of AI4REALNET is to develop an overarching multidisciplinary approach and to test and evaluate AI in industry-relevant use cases. The project team will combine emerging AI algorithms, existing open-source AI-friendly digital environments, socio-technical design of AI-based decision systems and human-machine interaction (HMI) to improve the operation of network infrastructures in real-time and predictive mode. The research aspects will be developed along three critical infrastructures whose virtual and physical assets, systems and networks are considered vital in Europe and whose disruption would have a crippling effect on society. These infrastructures come from the energy (electricity grid) and mobility (rail and air traffic management) sectors, two of the five priority sectors identified in the European national AI strategies. The project partners therefore include railroad companies such as SBB and Deutsche Bahn as well as air traffic control services in various countries.

Better decision-making through human-AI collaboration

The mutual complementarity of humans and AI offers many opportunities. For example, artificial intelligence can compensate for human weaknesses, and vice versa. For this approach to work, technical innovations should also be conceived from a human perspective. "Particularly in interdisciplinary projects, the technical perspective of applied psychology is relevant from the outset in order to develop solutions that are practicable and feasible for people in the long term," says Toni Wäfler. For Ricardo Chavarriaga, the vision of the AI4REALNET project is to explore the coexistence of human control and AI-based automation at different levels - from full human control with AI support to co-learning and trustworthy AI-based control. "For safety-critical applications, AI systems are traditionally trained, tested and then frozen so that they cannot change. We want to develop systems that can adapt and improve over time. The inclusion of the psychological perspective is essential in order to understand how lasting added value can be created from the interaction between people, technology and organization. The project therefore aims to empower people to improve their performance and achieve a higher level of reliability and security of critical infrastructures."