Dvij Kalaria

Associate Consultant

Berkeley, California, United States6 yrs 7 mos experience

Most Likely To SwitchAI Enabled

Key Highlights

Led vehicle dynamics team for Indy Autonomous Challenge.
Developed advanced control systems for autonomous racing.
Expert in robotics and machine learning applications.

Stackforce AI infers this person is a Robotics and Autonomous Systems expert with a focus on Machine Learning and Vehicle Dynamics.

Contact

dvij.kalaria@gmail.com LinkedIn

Skills

Core Skills

RoboticsComputer VisionVehicle DynamicsControlsMachine LearningRoad RacingDeep Learning

Other Skills

3D Reconstruction3d visionActive LearningAlgorithmsAndroid DevelopmentArtificial Intelligence (AI)Autonomous VehiclesC (Programming Language)C++CUDAGame DesignGame DevelopmentImage ProcessingIndustrial RoboticsMicrosoft Office

About

Hi, I am currently a masters in robotics student at CMU working with Dr. John Dolan and Dr. Qin Lin. Throughout my Master’s, my prime goal has been to develop robots that are 1) robust to sensor errors 2) non-conservative, 3) capable of using expert demonstrations and 4) of adapting to unknown situations safely. I am also he vehicle dynamics team lead of the AIRacingTech team for the Indy Autonomous Challenge (IAC). As a hobby, I love developing games, hiking, playing any racket sport in my free time

Experience

Carnegie mellon university robotics institute

2 roles

Teaching Assistant

Aug 2023 – Present · 2 yrs 7 mos · Pittsburgh, Pennsylvania, United States

TA for F1Tenth course for spring 2024 and fall 2023

Research Assistant

Jul 2021 – Present · 4 yrs 8 mos · Pittsburgh, Pennsylvania, United States

Work with CMU AV research lab advised by Prof. John Dolan and Prof. Qin Lin

Computer VisionDeep LearningPyTorchResearchRobot Operating System (ROS)TensorFlow+2

Ai racing tech

Student researcher, Vehicle dynamics team lead

Jul 2023 – Feb 2024 · 7 mos · Pittsburgh, Pennsylvania, United States · On-site

Implemented Model Predictive Control (MPC) control with a more complex vehicle model suitable at high speeds to consider objectives for overtaking, optimal racing line keeping and use of drafting. 
Implemented Adaptive LQG control followed by a CBF layer that considers lateral tire dynamics
Implemented online parameter estimator that can adapt controllers to friction, aerodynamic params
Implemented CBF for track boundary violations and safe distance keeping from opponent
Various engineering tasks like offline torque map generation, implementing ABS, developing numeric
simulator, parallel processing, bank angle estimation, sensor data filtering and bias corrections etc
AI Racing Tech team finished 3rd on the Indy Autonomous Challenge, CES 2023 at Las Vegas

C++Software DevelopmentVehicle DynamicsControlsMachine Learning

Pratt miller

Research Intern

May 2023 – Aug 2023 · 3 mos · Huntersville, North Carolina, United States · Hybrid

Worked under race strategy team.
Proposed ML models for opponents to predict their pit strategy, driver and vehicle limit parameters
Developed an optimization framework to get an optimal pit stop choice given the opponent ML models
Used RL to optimize ego agent race strategy against other trained racing agents based off previous races
Deployed the model on flask as a demo during a live IMSA race

Road RacingReinforcement LearningIndustrial RoboticsMachine Learning