Astha Gupta

Hello! I’m Astha Gupta, a robotics engineer and PhD student at EPFL, Switzerland. My research interests lie at the intersection of bio-inspired control, reinforcement learning, and robot locomotion. I am driven by a passion for understanding natural movement patterns to design robust and fault-tolerant controllers in robotic systems. My work focuses on leveraging deep reinforcement learning for sensory feedback design and sim-to-real transfer.

I specialize in robotic systems that are bio-inspired, focusing on amphibious robots modeled after eels and salamanders. The robots I have worked with include the eel-like AganathaX, the salamander-like Krock, and Pleurobot. My research aims to understand sensorimotor integration to achieve robust locomotion in challenging environments, with practical applications in search and rescue missions. During my master’s thesis, I designed and characterized multimodal controllers to replicate salamander locomotion using somatosensory feedback with Central Pattern Generators (CPGs). As part of my valorization project at EPFL’s Biorobotics Laboratory, I integrated the Krock robot with the PyBullet simulation platform, adapted its low-level controllers, and demonstrated simulation-to-real transfer for CPG-based locomotion strategies.

Educationally, I hold a European Master in Advanced Robotics Plus (EMARO+) from Ecole Centrale de Nantes and the University of Genoa, and a Master of Science in Information Systems from BITS Pilani. These experiences have equipped me with a robust technical background and the ability to tackle complex problems in robotics and AI.

In my professional journey, I have had the opportunity to work on fascinating projects, from developing controllers for amphibious robots to implementing 3D frameworks for spatio-temporal analysis of human-object interactions. My technical skills include programming in C, C++, and Python, as well as working with ROS, OpenCV, Keras, Tensorflow, OpenAI Gym and various machine learning libraries.

Beyond my work in robotics, I have strong skills in core computer science. At VMware, I focused on data center management and infrastructure, while at Inria, I implemented PAC Bayesian algorithms for data analysis. At iLabs [24]7 Inc., I developed data science projects involving advertisement re-targeting and product recommendation systems. Additionally, I am proficient in simulation frameworks like Simulink, Mujoco, Pybullet, Gazebo and have experience with single-board computers and microcontrollers such as RaspberryPi, Odroid-XU4, Arduino, and Nvidia Jetson Xavier.

I am always excited to discuss innovative ideas and potential collaborations. Feel free to reach out!