Hi! I am Siddharth!

I am a graduate student at University of Pennsylvania in the Electrical and Systems Engineering department . I work in the domain of General Robotics and Automation.

At Penn, I work at GRASP lab in the group of Professor Kostas Daniilidis under the mentorship of Bernadette Bucher. At GRASP I am working towards curiosity based reinforcement learning models and my current project involves the integration of curiosity in multi agent exploration and planning.

At Penn, I also work at MLab in the group of Professor Rahul Mangharam under the mentorship of Matthew O' Kelly. At Mlabs I am working towards the development of perception pipeline for autonomous vehicles which are tightly integrated with the path planning algorithms using a bottom up approach.

In the past, I have also worked as a motion path planning intern at Robotics Research Center, Hyderabad working towards the development of full scale autonomous vehicles reaching level-3 autonomy.

Email  /  Github  /  CV  /  Google Scholar  /  LinkedIn

profile photo
Work and Research

I have a few projects which span from fabricating single seater all terrain vehicles to making full scale autonomous vehicles move on their own using sensors. The projects are outlined below and if you are interested and/or want to know more about any of those, shoot me a mail!.

F1/10th Autonomous Racing: Autonomous vehicle development on 1/10th size platforms. The F1/10 website,

The project involves developing 1/10th scale autonomous vehicles with the complete AV stack. Includes perception, planning and control tightly bonded together.

Gradient Aware-Shrinking Domain based Control Design for Reactive Planning Frameworks used in Autonomous Vehicles. arvix link, video link,

As a part of the Autnomous Vehicle development project in collaboration with Mathworks and Intel at Robotics Research Center, IIIT Hyderabad. Robust frameworks for speed control using discrete time model predictive control using dynamic bicycle model and lateral approximations.

Autonomous Vehicle Development for Mahindra Rise Challanege RnD showcase video link, Drive-by-wire video link,

As a part of the Autnomous Vehicle development project in collaboration with Mathworks and Intel at Robotics Research Center, IIIT Hyderabad. Robust frameworks for speed control using discrete time model predictive control using dynamic bicycle model and lateral approximations.

Curiosity based reinforcement learning models for multi agent exploration and planning Project Link,

Part of the Honda Curious Minded Machines project. Exploring ways to make robots more curious. My particular work is in the domain of multi agent exploration and foraging. Trying to make agents more curious for exploration and faster convergence. Exploring ways if agents can learn to be curious.

RoboNet: Learning and Generalizing across Robots through Large-Scale Visual Prediction and Planning. Project Link, Paper Link,

Part of an inter-institutional project involving development of video prediction models which can be deployed on real robots to perform tasks. The prominent idea in this project is generalization of learning across robots using a huge multi-institutional data set which can be used with new( unseen and varied) robots via few shot and zero shot learning.

SAE BAJA : Building All Terrain Vehicles. website Link, Design Report

A national racing event of all terrain vehicles for designing and demonstrating capabilities of all terrain vehicles. This project is a odd one out in the list projects here solely because I have always found this very cool and still gives me a rush.

Flapping Flight Analysis of Birds github Link, Design Report

Part of senior design project. Involved the analysis of the wings of birds during the various phases of flapping flight to analyze the regions of high vibration using rolling shutter effect of camers.

Just Some Other Things I like to do

I relish food I create.

In my spare time I also like to write when I don't read. No, I am not publishing anytime soon. If I do it is gonna pop up here.