Sankalp Arora

I am a Masters student at the Robotics Institute (RI), School of Computer Science, Carnegie Mellon University, since Fall 2012. I work with Prof. Sanjiv Singh and Dr. Sebastian Scherer. I graduated in 2010 from the Delhi College Of Engineering, Delhi University, with a Bachelors degree in Electronics and Communication Engineering. I worked as a research associate at Field Robotics Center, RI from 2010-2012.

I am interested in developing robots that can operate safely in uncertain environments without compromising performance. I believe the key to designing such robots is making them adaptive to dynamic environments through a tight coupling between planning, perception and control. My research interest lies in designing algorithms that enable systems to represent changing environments, model the effects of the changes on the system and adapt their behavior.

The need for systems that adapt is highlighted by autonomous vehicles operating in unstructured or uncertain environments. For example, flying vehicles operating at low altitudes or autonomous cars moving through varying weather conditions need to respond quickly to the variations. Most systems do not model these phenomena, and end up making conservative assumptions about the perceptual and dynamic limits. The conservative limits are a result of using static models and constraints to account for worst case operating environments. Adapting the models online using data collected during its lifetime promises to overcome these issues. Use of adaptive models and motion planning while efficiently utilizing perceptual resources, will allow the systems to operate safely at their true limits. There is a strong need for such improved levels of autonomy for search and rescue, cargo delivery and medical evacuation, where human lives depend on robots moving quickly and safely in dynamic environments.

Email: asankalp AT cmu DOT edu

Recent Events