Stefano closes the gap between simulated and real software tests.
I am a researcher and Ph.D. student in the Automated Software Testing (AST) competence field at fortiss. My primary focus lies in the intricate realm of testing autonomous driving systems (ADS). Within this dynamic field, I concentrate on addressing and bridging the gap that often exists between simulated and real-world testing of ADS. Specifically, my work delves into the challenging domain of measuring and mitigating the disparities that arise during the evaluation of autonomous car systems based on artificial intelligence. Testing these systems in the real world is imperative to gauge their actual performance. However, employing real vehicles for testing can be prohibitively expensive and sometimes pose safety risks. To overcome these challenges, the industry widely employs simulators for testing.
While simulators offer a cost-effective and safer alternative, they inherently present limitations as they remain approximations of reality. Simulators simplify the rules of physics, interactions with the real world, and even the visual representation of the environment. This introduces discrepancies in the observations of vehicle behavior, as well as how the vehicle's cameras and sensors perceive the world. My research endeavors revolve around the critical mission of reducing these disparities. I want to enhance the authenticity of simulated tests, ensuring that they more accurately reflect real-world scenarios. By narrowing the gap between simulated and real-world testing, my work contributes to the refinement and reliability of autonomous driving systems, ultimately advancing the field toward safer and more efficient solutions.
My research focus is in the area of human-AI interaction. This is about designing AI systems in a way that people can work well with them. This means, for example, that AI systems must be human-centered and user-friendly to enable successful collaboration. Contrary to the impression given in many product demos, AI systems are not perfect but can produce errors. Particularly important is also My main focus is therefore on ensuring that users of AI systems are designed in such a way that their users are not affected by erroneous AI behavior. Especially in high-stakes applications such as aviation, it is essential that AI systems work reliably and can be quickly and safely taken over by human operators in case of errors. Human operators are always aware of the limitations of AI systems and can effectively contribute their expertise when the AI reaches its limits. A central aspect of my research is therefore the development of methods and technologies to increase the safety and reliability of AI systems. This also involves integrating human expertise into the development process of AI systems to enable the best possible collaboration between humans and AI. The starting point of my research is therefore always people: How do people work? How do they make decisions in a particular environment? What challenges do they face? These insights form the basis for exploring how AI fits into human work processes, taking into account its limitations. Overall, my goal is to improve human-AI interaction and help users of AI systems to realize their full potential without being affected by erroneous AI behavior that humans are always in full control, and that human capabilities are enhanced, not replaced, by AI.
What drives me is my curiosity and the desire to delve deeper into the intricacies of my field. The creative process, from conceptualization to realization, is immensely fulfilling. There's an inherent joy in bringing something entirely new and exciting into existence, something that has the potential to make a meaningful impact.
My overarching vision is transforming ADS into safe, reliable, and widely adopted solutions. The potential of automated driving to revolutionize our relationship with vehicles is a driving force behind my research. In practical terms, I envision the seamless integration of my research results into the fabric of autonomous driving technology. I aspire to contribute to a future where these systems are not only cutting-edge but also synonymous with safety and dependability.
The most significant challenge lies in the competitive and rigorous nature of the academic landscape. The challenge, therefore, is to demonstrate the intrinsic value of my work in this intellectual competition. It's an endeavor that demands not only depth of knowledge but also the ability to articulate and substantiate the significance of my contributions to a broader audience. Convincing others of the worthiness of my research can be an intricate and demanding process.
What keeps me motivated amidst these challenges is the recognition that, once proven and accepted by prestigious international conferences, my ideas, and solutions, no matter how incremental, have made a tangible impact. The realization that my work has reached a level where it is acknowledged by peers and experts in the field serves as a powerful motivator.
I have succeeded in having my paper accepted at ICST (IEEE International Conference on Software Testing, Verification and Validation). The paper deals with my research topic, in particular the application of generative neural networks in the creation of simulated scenarios that closely resemble real-life situations. This success not only confirms the value of my research but also contributes to the broader academic discourse in my field.
I've been engaged in programming since my elementary school days when my grandpa introduced me to creating simple scripts for controlling model trains. This initial exposure ignited my interest, and I continued experimenting with C programming until I formally pursued a bachelor’s in informatics. From that point onwards, I've been immersed in programming for nearly two decades.
What captivates me about programming is the ability to witness my ideas materialize and transform them into functional products or objects that can be immediately utilized. It's the fascinating process of turning conceptualizations into tangible, operational outcomes that makes programming an immensely rewarding skill to master. The instant gratification of seeing your creations come to life and the practical impact it can have on various domains are what continually drives my passion for programming.
I discovered fortiss through a meeting with one of the fortiss professors, who is now my supervisor. Our paths crossed during a self-driving vehicle hackathon in Switzerland, where he shared insights into the captivating research conducted at fortiss. Intrigued by the cutting-edge work and the prospect of contributing to it, I applied soon after graduating, eager to be a part of this innovative and dynamic research environment.
The working environment at fortiss is both relaxed and open to dialogue, creating a conducive atmosphere for creativity and collaboration. I appreciate the freedom to express my ideas and pursue topics I find promising, allowing me to study and research in areas I believe hold potential. One of the valuable aspects of fortiss is the opportunity to engage in discussions about my field with experts from diverse domains within fortiss, fostering a rich exchange of insights and perspectives.
My colleagues at fortiss exemplify a diverse yet harmonious dynamic. Despite hailing from varied cultures and backgrounds, we share an open and understanding relationship. The collaborative spirit is evident as we frequently assist each other in overcoming challenges and solving problems encountered in our respective research endeavors. This supportive atmosphere not only enhances our collective productivity but also fosters a sense of camaraderie, creating a positive and collaborative work environment.
Working at fortiss offers a unique advantage due to its open and diverse nature. The institute facilitates collaboration across multiple competence fields, encouraging a comprehensive approach to problem-solving. The collegial atmosphere fosters familiarity and open communication among colleagues.
I can only advise potential applicants to be convinced of their interests, be open to new ideas and believe in themselves. People often doubt whether they have the knowledge or skills to achieve their goals, but self-confidence is crucial. With an openness to personal and professional growth and confidence in your abilities, the supportive environment at fortiss is ideal for pursuing your goals with confidence.
![[Translate to English:] Anführungszeichen Visual](/fileadmin/_processed_/7/4/csm_anfuehrungszeichen_03b3788b37.png)
