Entrepreneur | Computer Scientist | Roboticist
I am the CTO & Co-founder of Scaled Robotics. I'm generally interested in leveraging movement and interaction to solve complex real-world perception problems. For more than a decade I worked on various problems in autonomy for ground, aerial and humanoid robots. Now I have focused my attention on utilizing AI, Robotics & ML to optimize construction.
Prior to founding Scaled Robotics, I was lucky to spend more than a decade working as a robotics researcher at some of the world's leading robotics labs. My alma mater includes the Computational Learning and Motor Control Laboratory (USC), the Max Planck Institute's Autonomous Motion Department, the GRASP Laboratory (UPenn), and the Robotics Institute (CMU) just to name a few.
Motion & Perception
There's an interesting theory in neuroscience that claims movement is fundamentally tied to intelligence. This is also supported by evidence in the evolutionary biology from single-cell organisms. Building on these insights, I'm interested in endowing autonomous systems the ability to learn and reason about their environment through movement and interaction. This also requires them to build appropriate environmental representations that can facilitate these actions. These abilities can ultimately allow these autonomous systems to operate in complex human environments with minimal to no supervision. You can have a look at some of my current and past projects in these areas below.
VISION BASED NAVIGATION
Vision contributes to nearly 80% of human perception and cognition. Endowing autonomous systems with the appropriate visual sensors can allow them to geometrically and semantically parse their environments building navigable representations. My research in this area has centered around navigation in novel and dynamic environments.
SHARED AUTONOMY IN COMPLEX ENVIRONMENTS
Robotic hardware is still limited by poor sensing and imprecise actuation. Hence executing robust robotic behaviors has been a challenge in complex environments. To build reliable real-world applications, coupling partial autonomy with (human) supervisory control can result in repeatable robotic processes.
ROBUST 3D RECOGNITION
Recognizing salient objects and estimating their pose accurately, can be tremendously beneficial to autonomous systems modeling their environment. This allows them to solve complex problems ranging from SLAM to Active Perception. I spent a significant portion of my career studying the problem of robust 3D recognition under heavy occlusion and clutter.
ACTIVE AND INTERACTIVE PERCEPTION
Systems capable of moving and (physically) interacting with their environment can induce both affine and non-affine transformations in them. The core of my research focuses on exploiting these abilities to reason and learn representations with limited or no prior data in novel unstructured environments.