QUT's Tobias Fischer Pioneers Robotics Research
Dr. Tobias Fischer's interest in robotics began with a high school project in Germany where he built a robot designed to avoid obstacles, marking a significant early inspiration in his career.
"It was a real 'wow' moment for the keen seventeen-year-old," Dr. Fischer said. From a plexiglass robot in Germany to advanced robotics research in Australia, Tobias's career path reflects his commitment to advancing robotics for complex, real-world applications.
"I found it fascinating how so many things needed to come together for a seemingly simple task such as avoiding obstacles," Dr. Fischer explained.
Despite initial setbacks, where the robot either crashed into objects or didn't move, these early experiences taught him valuable lessons in troubleshooting and persistence.
On the day of the final high school demonstration, despite its issues, Tobias eagerly anticipated showing off his robot. He placed the robot on the table, toggled the control, and to his great surprise, it worked! The robot successfully avoided all objects in its path, earning him the highest mark for his project.
"I still think back to this demonstration when my colleagues are plagued with the 'demo effect': the exact opposite where everything works fine until the moment of truth," Dr. Fischer said. "Max Born, a Nobel Prize-winning physicist, once compared science to navigating a jungle through trial and error-a concept that resonated with me early in my career."
As a lecturer at the QUT School of Electrical Engineering and Robotics, Dr. Fischer instills the importance of trial and error, learning from mistakes, and perseverance in his students.
Robots with human-like perception
Dr. Fischer's career focuses on the overarching research goal of equipping robots with the perceptual abilities necessary to safely and intelligently interact with humans.
"In pursuing this goal, my research has spanned the fields of robotics, computer vision, and artificial intelligence," Dr. Fischer said. "My research goes beyond technology-it aims to positively impact society. Equipping robots with human-like perception could transform areas like search and rescue, disaster relief, and space exploration, with broad implications for everyday life."
"My colleague at the QUT Centre for Robotics, Distinguished Professor Emeritus Peter Corke, frames our field's diversity with a comprehensive list of essential topics from mechanics to computer vision. His expertise across these areas is rare and inspiring. This diversity of topics that people are working on makes robotics so fascinating to me."
After high school in Germany, Dr. Fischer chose to study computer engineering at university to cover the various disciplines involved in intelligent robotics.
"I didn't particularly enjoy the electrical engineering part of it, but I loved software engineering, algorithms, programming, and neural networks," Dr. Fischer said. "This was also the time when I realized that despite studying toward a bachelor's degree, I wouldn't nearly be able to acquire all the knowledge that I wanted."
A game changer
Following a fortunate encounter, he connected with Distinguished Professor John Tsotsos from York University, Canada. "Within a week, we had agreed that I could visit Professor Tsotsos for nine months to write my bachelor's thesis," Dr. Fischer said.
Deeply researching a topic under structured guidance without being explicitly told what to do was a game changer. The thesis introduced him to many concepts he still researches today, such as using knowledge about the human visual system in robotics.
After his time in Canada, Dr. Fischer continued his path into robotics and robotic navigation, enrolling in a Master in Artificial Intelligence (AI) at the University of Edinburgh, which had a long history in AI.
"I was deeply interested in courses related to vision, covering multiple disciplines such as neuroscience, psychology, and robotics," he said. "In retrospect, my thesis was the first time that I made a (albeit minor) contribution to human knowledge. My advisor Jim Bednar was inspirational and very hands-on, which differed greatly from what I had previously encountered."
During his PhD studies at the prestigious Imperial College London, Dr. Fischer pioneered a transformative engineering approach that combines concepts from computer science and engineering with biology and social sciences.
"Specifically, I made advances in perceptual perspective taking, where a robot learns to infer how the visual scene is seen from the visual perspective of another agent, human or robot."
It was also during his PhD studies that he realized how poorly computer vision algorithms - that often work perfectly on constrained datasets - generalize to real-world settings when deploying them on a robot.
"My research on eye gaze and blink estimation has been highly impactful, advancing both theory and practical applications, for example, controlling a wheelchair purely based on gaze input."
His thesis was awarded the Queen Mary UK Best Thesis in Robotics Award 2018 and the Eryl Cadwaladr Davies Prize for the best thesis in the Electrical and Electronic Engineering Department at Imperial College 2018.
A move to Brisbane
After completing his PhD and one year of post-doc at Imperial College, Dr. Fischer sought new challenges and connected with Professor Michael Milford, now Director of the QUT Centre for Robotics. This led to his move to Brisbane just before COVID struck.
"Michael's research aligned closely with mine. Additionally, the then Australian Centre for Robotic Vision based at QUT gave Australian research in robotic vision a huge push forward," Dr. Fischer said. "The ACRV, and the QUT Centre of Robotics that followed it, have raised global awareness of Australia as a leader in robotics. Being led out of QUT, I knew that there would be many more world-leading robotics researchers."
"At our QUT Centre for Robotics, we are at the forefront of global robotics research, offering a unique window into projects from underwater robots to drones and space robotics-a diversity that sets us apart."