Our study introduces a novel methodology and integrated system to address odour-based navigation challenges. We tackle this complex task by adopting a line-following approach. Drawing inspiration from observed insect behaviour, we employ a stereo olfaction technique, utilizing our custom made nano-material-based sensors. The resulting sensor data is channelled into our bio-inspired neural network, initially hosted on a Jetson and planned for transition onto the SpiNNaker-2 platform. This integration is effectively implemented on a 4-wheel rover, with future prospects for deployment on a quadruped robot. Our research contributes significantly to the progression of odour-based navigation solutions.
Rebecca Miko began her PhD in 2018 to contribute towards solving odour-based navigation in robotics, using neuromorphic hardware and event-based methods, after completing her MSc in Artificial Intelligence with Robotic specialism in 2017. She is now a research assistant (WHK) in the Chair of Materials Science and Nanotechnology at TU Dresden working as part of the 6G-life project. Her work involves integrating robotic systems with graphene nano sensors and neuromorphic hardware. For the purposes of odour-based navigation, she creates computationally efficient neural networks which extract features from sensory signals using event-based techniques and translate this into efficient robotic navigation algorithms.
Our study introduces a novel methodology and integrated system to address odour-based navigation challenges. We tackle this complex task by adopting a line-following approach. Drawing inspiration from observed insect behaviour, we employ a stereo olfaction technique, utilizing our custom made nano-material-based sensors. The resulting sensor data is channelled into our bio-inspired neural network, initially hosted on a Jetson and planned for transition onto the SpiNNaker-2 platform. This integration is effectively implemented on a 4-wheel rover, with future prospects for deployment on a quadruped robot. Our research contributes significantly to the progression of odour-based navigation solutions.
Rebecca Miko began her PhD in 2018 to contribute towards solving odour-based navigation in robotics, using neuromorphic hardware and event-based methods, after completing her MSc in Artificial Intelligence with Robotic specialism in 2017. She is now a research assistant (WHK) in the Chair of Materials Science and Nanotechnology at TU Dresden working as part of the 6G-life project. Her work involves integrating robotic systems with graphene nano sensors and neuromorphic hardware. For the purposes of odour-based navigation, she creates computationally efficient neural networks which extract features from sensory signals using event-based techniques and translate this into efficient robotic navigation algorithms.
