4 papers · 1 filter
Learning the Right Abstraction: Neural Reduced Dynamics for Complex Robot Control
Harry Zhang, Dan Negrut
High-fidelity embodied AI simulators provide realistic evaluation of complex robotic systems, but their computational cost limits their direct use for large-scale reinforcement lea…
Quantifying the Sim2real Gap for GPS and IMU Sensors
Ishaan Mahajan, Huzaifa Unjhawala, Harry Zhang +7
Simulation can and should play a critical role in the development and testing of algorithms for autonomous agents. What might reduce its impact is the ``sim2real'' gap -- the algor…
Zero-Shot Policy Transferability for the Control of a Scale Autonomous Vehicle
Harry Zhang, Stefan Caldararu, Sriram Ashokkumar +6
We report on a study that employs an in-house developed simulation infrastructure to accomplish zero shot policy transferability for a control policy associated with a scale autono…
Using simulation to design an MPC policy for field navigation using GPS sensing
Harry Zhang, Stefan Caldararu, Ishaan Mahajan +8
Modeling a robust control system with a precise GPS-based state estimation capability in simulation can be useful in field navigation applications as it allows for testing and vali…