We study sophisticated motor behaviors and their neural implementations for bio-inspired engineering.

Animals including human invest significant neuronal and metabolic resources for motor tasks. While the higher cognitive circuits responsible for motor-learning and the exact biomechanics underlying the behavior can vary greatly across animals, the sensorimotor strategies and neuromechanics that enable complex behaviors can transcend the individuality of animal species and even robots. I believe that this is the interface where we can take the first principles from the biological systems to bridge our rapidly developing artificial intelligence with advanced mechatronics.


1) Mechanosensation of flight & aerial interactions


Mechanosensation is a crucial sensory modality for motor-control. As we walk, we feel the ground to adjust our gaits, stagger to save a misstep, and catch ourselves when we trip. Similarly, for flying animals, sensing the air is like feeling the ground for us. Fluid sensing is a form of mechanosensation that is largely understudied, partly because fluid is difficult to characterize and mechanosensors are often inaccessible for in vivo measurements. Flying insects offer a great system for understanding the neural representation of aerodynamics. With our recent success of in-flight neural recording via an ultralight wireless backpack on the dragonfly, we can start to eavesdrop on mechanosensory signals during insect flight. This research has direct implications for the emerging fly-by-feel control (flight control using mechanosensory data) for aerial robots and also sets the stage for studying aerial interactions in animals and machines.

2) The neuromechanics of visual tracking, guidance and navigation


Vision is realized through action. In fact, most visual animals perform saccadic eye movements, smooth target tracking, and involuntary visual fixation. These visual behaviors have functional implications for visual acuity, motion vision, visual guidance and depth estimation. From the bioengineering standpoint, we are interested in the functions and implementation of visual behaviors. Flying insects perform a full repertoire of visual behaviors from target tracking, obstacle negotiation to navigation. We have developed an insect-scale motion capture protocol which can track the 3D kinematics of a dragonfly’s head, body, and wings. This approach allows us to reconstruct the visual gaze, body states, and wing dynamics from different freely behaving insects. Understanding the heuristics and the neural implementation underlying these visual behaviors can help us develop bio-inspired machine vision systems which incorporate appropriate motor gestures.

3) Neural implants and bio-telemetry technologies


Arguably the biggest challenge in modern microelectronics is wiring. We have a variety of implantable sensor or electrode arrays with high channel-count, yet we do not have any easy way to acquire the data without complicated and sometimes prohibitive wiring. Recent wireless neural implants have demonstrated the potential solution to this “wiring issue”. However, there are still many technical challenges that need to be overcome and many ideas to be tested. We will use large insects as a platform to experiment novel neural technologies. Insects’ distributed nervous system allows us to target different types of physiological signals and the ease of working with insects allows us to go through many design iterations quickly.

4) Soft robotics in the context of dynamic morphing


What is soft? We describe something as “soft” when it undergoes large deformation during normal operation. In that sense, most flying insects are soft because their wings deform dramatically under aerodynamic and inertial loads. Conversely, many soft-bodied animals can maintain specific body postures dynamically. At high-speed, some statically rigid systems become soft and some soft systems become relatively rigid. This is a fascinating yet relatively unexplored topic in biomechanics. Understanding dynamic morphing has vast implications for the development of soft robotics.