Christoph Vernaleken is a physics graduate of the University of Würzburg. Subsequently, he obtained a PhD in engineering from Darmstadt University of Technology, where he spent several years as a research associate working on flight deck and pilot assistance functions aimed at improving flight safety. After almost 15 years with the Human Factors Engineering team at Airbus Defence and Space in Manching, Germany, covering a broad range of subjects ranging from air-to-air refueling on the A400M to the application of VR in air traffic control, a flight deck for suborbital aircraft and optimized satellite ground stations, he now has a leading role in the development of the new flight deck for the D328eco regional turboprop aircraft as Senior Expert Flight Deck and Human Factors with Deutsche Aircraft GmbH in Oberpfaffenhofen.
Since 2020, Christoph has also been a lecturer at TU München, giving a course on Human Factors in aviation at the Chair of Ergonomics. Apart from various papers and textbook contributions in the domain of Human Factors and flight deck development, he has also published several articles and co-authored a book in the field of aviation history.
Human Factors Challenges Resulting From Automation and Artificial Intelligence
In the domain of Human Factors it has been known for a long time that complex automated systems may present unique challenges for human-machine interaction if users do not have sufficient awareness of the various automation modes the system provides, the conditions for transitions between these modes, and the respective functional capabilities an limitations associated with each mode. This may lead to so-called “automation surprises”, which often bring human-machine interaction almost or actually to a breakdown when users or operators are no longer able to understand the current behavior of the machine, let alone to predict its future actions. While the effects of such a breakdown might be limited to mere nuisance many everyday applications, such as word processing, there may be dire consequences in safety-critical applications. In aviation, for example, this has already led to several tragical accidents and many incidents in the past. With this in mind, this talk attempts to give some first insights in the specific new challenges that might arise for human-machine interfaces when “classic” algorithmic automation is increasingly superseded by automaton driven by artificial intelligence technologies, and why “trustworthy AI” is potentially not the best objective from either an engineering or a human factors perspective.
