Chinmay Vilas Samak

Tanmay Vilas Samak, Chinmay Vilas Samak, Venkat Krovi

This work presents a modular and parallelizable multi-agent deep reinforcement learning framework for imbibing cooperative as well as competitive behaviors within autonomous vehicles. We introduce AutoDRIVE Ecosystem as an enabler to develop physically accurate and graphically realistic digital twins of Nigel and F1TENTH, two scaled autonomous vehicle platforms with unique qualities and capabilities, and leverage this ecosystem to train and deploy multi-agent reinforcement learning policies. We first investigate an intersection traversal problem using a set of cooperative vehicles (Nigel) that share limited state information with each other in single as well as multi-agent learning settings using a common policy approach. We then investigate an adversarial head-to-head autonomous racing problem using a different set of vehicles (F1TENTH) in a multi-agent learning setting using an individual policy approach. In either set of experiments, a decentralized learning architecture was adopted, which allowed robust training and testing of the approaches in stochastic environments, since the agents were mutually independent and exhibited asynchronous motion behavior. The problems were further aggravated by providing the agents with sparse observation spaces and requiring them to sample control commands that implicitly satisfied the imposed kinodynamic as well as safety constraints. The experimental results for both problem statements are reported in terms of quantitative metrics and qualitative remarks for training as well as deployment phases.

Chinmay Vilas Samak, Tanmay Vilas Samak, Venkat Narayan Krovi

Multi-agent reinforcement learning (MARL) for cyber-physical vehicle systems usually requires a significantly long training time due to their inherent complexity. Furthermore, deploying the trained policies in the real world demands a feature-rich environment along with multiple physical embodied agents, which may not be feasible due to monetary, physical, energy, or safety constraints. This work seeks to address these pain points by presenting a mixed-reality (MR) digital twin (DT) framework capable of: (i) boosting training speeds by selectively scaling parallelized simulation workloads on-demand, and (ii) immersing the MARL policies across hybrid simulation-to-reality (sim2real) experiments. The viability and performance of the proposed framework are highlighted through two representative use cases, which cover cooperative as well as competitive classes of MARL problems. We study the effect of: (i) agent and environment parallelization on training time, and (ii) systematic domain randomization on zero-shot sim2real transfer, across both case studies. Results indicate up to 76.3% reduction in training time with the proposed parallelization scheme and sim2real gap as low as 2.9% using the proposed deployment method.

Chinmay Vilas Samak, Tanmay Vilas Samak, Sivanathan Kandhasamy

In this work, we present a rigorous end-to-end control strategy for autonomous vehicles aimed at minimizing lap times in a time attack racing event. We also introduce AutoRACE Simulator developed as a part of this research project, which was employed to simulate accurate vehicular and environmental dynamics along with realistic audio-visual effects. We adopted a hybrid imitation-reinforcement learning architecture and crafted a novel reward function to train a deep neural network policy to drive (using imitation learning) and race (using reinforcement learning) a car autonomously in less than 20 hours. Deployment results were reported as a direct comparison of 10 autonomous laps against 100 manual laps by 10 different human players. The autonomous agent not only exhibited superior performance by gaining 0.96 seconds over the best manual lap, but it also dominated the human players by 1.46 seconds with regard to the mean lap time. This dominance could be justified in terms of better trajectory optimization and lower reaction time of the autonomous agent.

Chinmay Vilas Samak, Tanmay Vilas Samak, Sivanathan Kandhasamy

This work presents proximally optimal predictive control algorithm, which is essentially a model-based lateral controller for steered autonomous vehicles that selects an optimal steering command within the neighborhood of previous steering angle based on the predicted vehicle location. The proposed algorithm was formulated with an aim of overcoming the limitations associated with the existing control laws for autonomous steering - namely PID, Pure-Pursuit and Stanley controllers. Particularly, our approach was aimed at bridging the gap between tracking efficiency and computational cost, thereby ensuring effective path tracking in real-time. The effectiveness of our approach was investigated through a series of dynamic simulation experiments pertaining to autonomous path tracking, employing an adaptive control law for longitudinal motion control of the vehicle. We measured the latency of the proposed algorithm in order to comment on its real-time factor and validated our approach by comparing it against the established control laws in terms of both crosstrack and heading errors recorded throughout the respective path tracking simulations.

Tanmay Vilas Samak, Chinmay Vilas Samak, Ming Xie

AutoDRIVE is envisioned to be an integrated research and education platform for scaled autonomous vehicles and related applications. This work is a stepping-stone towards achieving the greater goal of realizing such a platform. Particularly, this work introduces the AutoDRIVE Simulator, a high-fidelity simulator for scaled autonomous vehicles. The proposed simulation ecosystem is developed atop the Unity game engine, and exploits its features in order to simulate realistic system dynamics and render photorealistic graphics. It comprises of a scaled vehicle model equipped with a comprehensive sensor suite for redundant perception, a set of actuators for constrained motion control and a fully functional lighting system for illumination and signaling. It also provides a modular environment development kit, which comprises of various environment modules that aid in reconfigurable construction of the scene. Additionally, the simulator features a communication bridge in order to extend an interface to the autonomous driving software stack developed independently by the users. This work describes some of the prominent components of this simulation system along with some key features that it has to offer in order to accelerate education and research aimed at autonomous driving.

Chinmay Vilas Samak, Tanmay Vilas Samak, Sivanathan Kandhasamy

This chapter focuses on the self-driving technology from a control perspective and investigates the control strategies used in autonomous vehicles and advanced driver-assistance systems from both theoretical and practical viewpoints. First, we introduce the self-driving technology as a whole, including perception, planning and control techniques required for accomplishing the challenging task of autonomous driving. We then dwell upon each of these operations to explain their role in the autonomous system architecture, with a prime focus on control strategies. The core portion of this chapter commences with detailed mathematical modeling of autonomous vehicles followed by a comprehensive discussion on control strategies. The chapter covers longitudinal as well as lateral control strategies for autonomous vehicles with coupled and de-coupled control schemes. We as well discuss some of the machine learning techniques applied to autonomous vehicle control task. Finally, we briefly summarize some of the research works that our team has carried out at the Autonomous Systems Lab and conclude the chapter with a few thoughtful remarks.

Sivanathan Kandhasamy, Vinayagam Babu Kuppusamy, Tanmay Vilas Samak et al.

This work presents a decentralized motion planning framework for addressing the task of multi-robot navigation using deep reinforcement learning. A custom simulator was developed in order to experimentally investigate the navigation problem of 4 cooperative non-holonomic robots sharing limited state information with each other in 3 different settings. The notion of decentralized motion planning with common and shared policy learning was adopted, which allowed robust training and testing of this approach in a stochastic environment since the agents were mutually independent and exhibited asynchronous motion behavior. The task was further aggravated by providing the agents with a sparse observation space and requiring them to generate continuous action commands so as to efficiently, yet safely navigate to their respective goal locations, while avoiding collisions with other dynamic peers and static obstacles at all times. The experimental results are reported in terms of quantitative measures and qualitative remarks for both training and deployment phases.

Tanmay Vilas Samak, Chinmay Vilas Samak, Sivanathan Kandhasamy

In this work, we present a lightweight pipeline for robust behavioral cloning of a human driver using end-to-end imitation learning. The proposed pipeline was employed to train and deploy three distinct driving behavior models onto a simulated vehicle. The training phase comprised of data collection, balancing, augmentation, preprocessing and training a neural network, following which, the trained model was deployed onto the ego vehicle to predict steering commands based on the feed from an onboard camera. A novel coupled control law was formulated to generate longitudinal control commands on-the-go based on the predicted steering angle and other parameters such as actual speed of the ego vehicle and the prescribed constraints for speed and steering. We analyzed computational efficiency of the pipeline and evaluated robustness of the trained models through exhaustive experimentation during the deployment phase. We also compared our approach against state-of-the-art implementation in order to comment on its validity.