This paper introduces SPIRAL (Self-Play Incremental Racing Algorithm for Learning), a novel approach for training autonomous drones in multi-agent racing competitions. SPIRAL distinctively employs a self-play mechanism to incrementally cultivate complex racing behaviors within a challenging, dynamic environment. Through this self-play core, drones continuously compete against increasingly proficient versions of themselves, naturally escalating the difficulty of competitive interactions. This progressive learning journey guides agents from mastering fundamental flight control to executing sophisticated cooperative multi-drone racing strategies. Our method is designed for versatility, allowing integration with any state-of-the-art Deep Reinforcement Learning (DRL) algorithms within its self-play framework. Simulations demonstrate the significant advantages of SPIRAL and benchmark the performance of various DRL algorithms operating within it. Consequently, we contribute a versatile, scalable, and self-improving learning framework to the field of autonomous drone racing. SPIRAL's capacity to autonomously generate appropriate and escalating challenges through its self-play dynamic offers a promising direction for developing robust and adaptive racing strategies in multi-agent environments. This research opens new avenues for enhancing the performance and reliability of autonomous racing drones in increasingly complex and competitive scenarios.

Generalization to unseen environments is a significant challenge in the field of robotics and control. In this work, we focus on contextual reinforcement learning, where agents act within environments with varying contexts, such as self-driving cars or quadrupedal robots that need to operate in different terrains or weather conditions than they were trained for. We tackle the critical task of generalizing to out-of-distribution (OOD) settings, without access to explicit context information at test time. Recent work has addressed this problem by training a context encoder and a history adaptation module in separate stages. While promising, this two-phase approach is cumbersome to implement and train. We simplify the methodology and introduce SP ARC: single-phase adaptation for robust control. We test SP ARC on varying contexts within the high-fidelity racing simulator Gran Turismo 7 and wind-perturbed MuJoCo environments, and find that it achieves reliable and robust OOD generalization.

Abstract-- Autonomous racing requires tight integration between perception, planning and control to minimize latency as well as timely decision making. A standard autonomy pipeline comprising of a global planner, local planner, and controller loses information as the higher-level racing context is sequentially propagated downstream into specific task-oriented context. In particular, the global planner's understanding of optimality is typically reduced to a sparse set of waypoints, leaving the local planner to make reactive decisions with limited context. This paper investigates whether additional global insights, specifically time-optimality information, can be meaningfully passed to the local planner to improve downstream decisions. We introduce a framework that preserves essential global knowledge and convey it to the local planner through QuayPoints - regions where deviations from the optimal raceline result in significant compromises to optimality. QuayPoints enable local planners to make more informed global decisions when deviating from the raceline, such as during strategic overtaking. T o demonstrate this, we integrate QuayPoints into an existing planner and show that it consistently overtakes opponents traveling at up to 75% of the ego vehicle's speed across four distinct race tracks.

Abstract--This study introduces a haptic shared control framework designed to teach human drivers advanced driving skills. In this context, shared control refers to a driving mode where the human driver collaborates with an autonomous driving system to control the steering of a vehicle simultaneously. Advanced driving skills are those necessary to safely push the vehicle to its handling limits in high-performance driving such as racing and emergency obstacle avoidance. Previous research has demonstrated the performance and safety benefits of shared control schemes using both subjective and objective evaluations. However, these schemes have not been assessed for their impact on skill acquisition on complex and demanding tasks. Prior research on long-term skill acquisition either applies haptic shared control to simple tasks or employs other feedback methods like visual and auditory aids. T o bridge this gap, this study creates a cyber racing coach framework based on the haptic shared control paradigm and evaluates its performance in helping human drivers acquire high-performance driving skills. The framework introduces (1) an autonomous driving system that is capable of cooperating with humans in a highly performant driving scenario; and (2) a haptic shared control mechanism along with a fading scheme to gradually reduce the steering assistance from autonomy based on the human driver's performance during training. Two benchmarks are considered: self-learning (no assistance) and full assistance during training. Results from a human subject study indicate that the proposed framework helps human drivers develop superior racing skills compared to the benchmarks, resulting in better performance and consistency. Advanced driving skills refer to a set of competencies that go beyond basic driving abilities in terms of situational awareness, hazard perception, risk management, and vehicle handling [1]. They are crucial in high-performance driving tasks such as racing, and can also improve safety in everyday driving [1], [2]. This work has been submitted to the IEEE for possible publication.

Autonomous driving applications have become increasingly relevant in the automotive industry due to their potential to enhance vehicle safety, efficiency, and user experience, thereby meeting the growing demand for sophisticated driving assistance features. However, the development of reliable and trustworthy autonomous systems poses challenges such as high complexity, prolonged training periods, and intrinsic levels of uncertainty. Mini Autonomous Cars (MACs) are used as a practical testbed, enabling validation of autonomous control methodologies on small-scale setups. This simplified and cost-effective environment facilitates rapid evaluation and comparison of machine learning models, which is particularly useful for algorithms requiring online training. To address these challenges, this work presents a methodology based on RGB-D information and three-dimensional convolutional neural networks (3D CNNs) for MAC autonomous driving in simulated environments. We evaluate the proposed approach against recurrent neural networks (RNNs), with architectures trained and tested on two simulated tracks with distinct environmental features. Performance was assessed using task completion success, lap-time metrics, and driving consistency. Results highlight how architectural modifications and track complexity influence the models' generalization capability and vehicle control performance. The proposed 3D CNN demonstrated promising results when compared with RNNs.

Controlling autonomous vehicles at their handling limits is a significant challenge, particularly for electric vehicles with active four wheel drive (A4WD) systems offering independent wheel torque control. While traditional Vehicle Dynamics Control (VDC) methods use complex physics-based models, this study explores Deep Reinforcement Learning (DRL) to develop a unified, high-performance controller. We employ the Proximal Policy Optimization (PPO) algorithm to train an agent for optimal lap times in a simulated racecar (TORCS) at the tire grip limit. Critically, the agent learns an end-to-end policy that directly maps vehicle states, like velocities, accelerations, and yaw rate, to a steering angle command and independent torque commands for each of the four wheels. This formulation bypasses conventional pedal inputs and explicit torque vectoring algorithms, allowing the agent to implicitly learn the A4WD control logic needed for maximizing performance and stability. Simulation results demonstrate the RL agent learns sophisticated strategies, dynamically optimizing wheel torque distribution corner-by-corner to enhance handling and mitigate the vehicle's inherent understeer. The learned behaviors mimic and, in aspects of grip utilization, potentially surpass traditional physics-based A4WD controllers while achieving competitive lap times. This research underscores DRL's potential to create adaptive control systems for complex vehicle dynamics, suggesting RL is a potent alternative for advancing autonomous driving in demanding, grip-limited scenarios for racing and road safety.

Autonomous racing presents unique challenges due to its non-linear dynamics, the high speed involved, and the critical need for real-time decision-making under dynamic and unpredictable conditions. Most traditional Reinforcement Learning (RL) approaches rely on extensive simulation-based pre-training, which faces crucial challenges in transfer effectively to real-world environments. This paper introduces a robust on-board RL framework for autonomous racing, designed to eliminate the dependency on simulation-based pre-training enabling direct real-world adaptation. The proposed system introduces a refined Soft Actor-Critic (SAC) algorithm, leveraging a residual RL structure to enhance classical controllers in real-time by integrating multi-step Temporal-Difference (TD) learning, an asynchronous training pipeline, and Heuristic Delayed Reward Adjustment (HDRA) to improve sample efficiency and training stability. The framework is validated through extensive experiments on the F1TENTH racing platform, where the residual RL controller consistently outperforms the baseline controllers and achieves up to an 11.5 % reduction in lap times compared to the State-of-the-Art (SotA) with only 20 min of training. Additionally, an End-to-End (E2E) RL controller trained without a baseline controller surpasses the previous best results with sustained on-track learning.

Scientific development often takes place in the context of research projects carried out by dedicated students during their time at university. In the field of self-driving software research, the Formula Student Driverless competitions are an excellent platform to promote research and attract young engineers. This article presents the software stack developed by BME Formula Racing Team, that formed the foundation of the development that ultimately led us to full-scale autonomous racing. The experience we gained here contributes greatly to our successful participation in the Abu Dhabi Autonomous Racing League. We therefore think it is important to share the system we used, providing a valuable starting point for other ambitious students. We provide a detailed description of the software pipeline we used, including a brief description of the hardware-software architecture. Furthermore, we introduce the methods that we developed for the modules that implement perception; localisation and mapping, planning, and control tasks.