Internship at Argonne National Lab
We asked our students who were interning over the summer to share there experience through a blog post. Here is what Shobhit Gupta, a graduate research assistant at CAR and PhD student in the Department of Mechanical Engineering has to say about his experience with Argonne National Laboratory.
Over the summer of 2021, I received the opportunity to pursue an internship in the Vehicle & Mobility Systems (VMS) group at Argonne National Laboratory. Argonne’s VMS group has been at the forefront of energy-efficient transportation with particular focus on Connected and Automated Vehicles (CAVs) and electric aviation and has made a huge impact in these areas with the development of cutting-edge simulation tools and packages namely RoadRunner (RR), Autonomie, SVTrip, etc. I worked with the CAV controls group within the VMS team that focuses on developing novel control techniques to improve the energy efficiency of CAVs (referred as “Eco-Driving”) and test them on real-world scenarios in RR.
During my internship, I was assigned a project to develop a data-driven powertrain-agnostic Eco-Driving speed planning algorithm/framework for CAVs. A majority of the current CAV research is focused on either ensuring safety through perception, localization and planning, or improving energy-efficiency. These activities require developing state-of-the-art control algorithms that can be deployed on vehicles equipped with Rapid Prototyping Systems (RPS) offering higher computational and memory resources. However, only a few percentages of the current vehicle fleets have these advanced RPS, therefore the penetration of the eco-driving technology is limited to only advanced future vehicles. This internship was an attempt to make possible the scalability of eco-driving controls to existing real-world controllers.
Prior to the internship, I had the experience of using model-based optimal control techniques but was entirely new to the area of data-driven controls. The challenging part was to first understand the theoretical concepts, combine them with novel control techniques to develop a “Fast Data-Driven Predictive Control Algorithm” to solve the Eco-Driving problem and then eventually create an automated workflow in RR to test real-world scenarios. I was fortunate to have worked with an amazing team and mentors that helped me achieve all the tasks in limited time. This work has laid a foundation for the team to develop data-driven controls and will soon be published in one of the automotive control conferences.
Due to COVID restrictions, the internship was off-site, but the team ensured that I contribute and remain engaged in various activities at Argonne. The most highlighting part of this internship was to attend the seminar where US Secretary of Energy, Jennifer M. Granholm addressed all the national lab interns about the impact an intern can make towards the topics related to climate change.
I am grateful to my advisor Dr. Marcello Canova who has over the years helped me grow both professionally and academically by inculcating in me the critical thinking required to solve a new challenging problem. I also find myself fortunate to have been working with the NEXTCAR team and great minds at the Center for Automotive Research whose diverse knowledge in different areas helped me succeed in this internship.
