View the Project on GitHub anandsingh3996/Anandsingh-Chauhan
Please see a summary of my experience and accomplishments below, you can also view my full CV.
You can contact me at: anandsingh.chauhan@alumni.iitgn.ac.in
I'm currently a Researcher at TCS Research in the Data and Decision Sciences Research Area. My work is fueled by a genuine passion for harnessing the potential of machine learning and operational research to address complex, real-world challenges. Throughout my career, I've been dedicated to bridging the gap between theory and practical solutions, and I bring with me a wide-ranging skill set that covers various domains. This includes my in-depth expertise in reinforcement learning and its diverse applications, ranging from robust power network control to optimizing supply chains, facilitating peer-to-peer energy trading, and remaining at the cutting edge of innovations in electric vehicle technology.
M.Tech. , Electrical Engineering, Indian Institute of Technology Gandhinagar (July 2018 – August 2020)
B.E., Electrical Engineering, L.D. College of Engineering (August 2014 – June 2018)
Researcher @ Data & Decision Sciences, Tata Consultancy Services Limited - Research (September 2020 - Present)
We develop a RL framework, PowRL, to mitigate the effects of unexpected network events, as well as reliably maintain electricity everywhere on the network at all times. The PowRL leverages a novel heuristic for overload management, along with the RL-guided decision making on optimal topology selection to ensure that the grid is operated safely and reliably (with no overloads). Even with its reduced action space, PowRL tops the leaderboard in the L2RPN NeurIPS 2020 challenge (Robustness track) at an aggregate level, while also being the top performing agent in the L2RPN WCCI 2020 challenge. he extension of this work focuses on the development of neural network based learning framework for the optimal generator dispatch and battery scheduling.
We develop an innovative algorithmic framework to optimize cost-to-serve (CTS) in e-commerce, addressing the challenge of efficiently fulfilling dynamically generated orders from multiple customers across various warehouses and vehicle fleets. This project incorporated a two-level decision-making process: firstly, selecting the optimal fulfillment node for each order (including deferral options), and secondly, routing vehicles efficiently to deliver orders from the same warehouse. Our approach combined graph neural networks, reinforcement learning, for node fulfillment and vehicle routing, while considering real-world constraints such as warehouse inventory capacity, vehicle characteristics, and customer delivery time windows.
We addressed the challenge of optimizing productivity in modern warehousing by developing a heuristic-guided Reinforcement Learning(RL) agent. The project aimed to minimize both robot travel distance and task execution delays while considering practical constraints like charging/discharging and collision-free navigation. The developed RL based framework outperformed industry-standard practices such as FIFO and a myopic greedy heuristic. This innovative approach enhances efficiency and operational agility in dynamic warehousing environments.
We developed a Reinforcement Learning(RL) based framework for optimizing the replenishment process of multiple products with varying lead times. The system efficiently forecasts upcoming demand while considering factors such as shelf life, lead times, and other critical variables. This project showcased the ability to apply advanced AI techniques to enhance manufacturing and supply chain operations.
Contributed to the establishment of a cutting-edge hardware and software platform for Peer-to-Peer (P2P) energy trading at IIT Gandhinagar. The project centered around creating a real-world testbed involving two prosumers (Peer A and Peer B) and a consumer (Peer C). Peer A utilized solar PV and battery energy storage, while Peer B featured electric vehicle (EV) charging capabilities with Vehicle-to-Grid (V2G) functionality. Additionally, the testbed seamlessly interfaced with a Blockchain-based digital platform to demonstrate Peer-to-Peer energy trading in a practical environment. This work aligns with the evolving energy sector’s transformation, facilitating the integration of renewables, operational efficiency improvement, and transactive energy dynamics.