Abstract:This research aims to investigate professional racing drivers' expertise to develop an understanding of their cognitive and adaptive skills to create new autonomy algorithms. An expert interview study was conducted with 11 professional race drivers, data analysts, and racing instructors from across prominent racing leagues. The interviews were conducted using an exploratory, non-standardized expert interview format guided by a set of prepared questions. The study investigates drivers' exploration strategies to reach their vehicle limits and contrasts them with the capabilities of state-of-the-art autonomous racing software stacks. Participants were questioned about the techniques and skills they have developed to quickly approach and maneuver at the vehicle limit, ultimately minimizing lap times. The analysis of the interviews was grounded in Mayring's qualitative content analysis framework, which facilitated the organization of the data into multiple categories and subcategories. Our findings create insights into human behavior regarding reaching a vehicle's limit and minimizing lap times. We conclude from the findings the development of new autonomy software modules that allow for more adaptive vehicle behavior. By emphasizing the distinct nuances between manual and autonomous driving techniques, the paper encourages further investigation into human drivers' strategies to maximize their vehicles' capabilities.

What problem does this paper attempt to address?

The paper mainly discusses how to develop autonomous driving algorithms by studying the driving skills of professional race car drivers, in order to improve the performance of autonomous vehicles in racing. The study conducted interviews with 11 professional race car drivers, data analysts, and racing coaches using expert interview method, analyzed how race car drivers explore and approach the limits of the vehicle, and compared them with the state-of-the-art autonomous driving software stack. The study found that human race car drivers can quickly improve lap times and approach the vehicle's performance potential through years of experience, sensory integration, and adaptive learning. In contrast, current autonomous driving systems still lag behind human drivers in some aspects. The paper proposes three main contributions: 1. Demonstrates the interview results of 11 professionals in racing. 2. Describes the strategies that human drivers use to explore the limits of the vehicle and compares them with the capabilities of autonomous driving software. 3. Extracts insights from the comparison and proposes to enhance the performance of modular autonomous driving software in racing based on the strategies of human drivers. The paper also reviews existing racing software architectures, including end-to-end machine learning and modular layered approaches, and points out the challenges of existing autonomous driving systems in terms of data requirements, safety, and environmental adaptability. The research methodology includes designing expert interviews, organizing data through qualitative content analysis framework, and summarizing multiple categories and subcategories. The results indicate that although autonomous driving systems perform well in certain environments, they have not yet reached the level of human drivers, especially in quickly approaching and manipulating the limits of the vehicle. The paper concludes by discussing how to learn from human driver behavior to improve the design of autonomous driving software, particularly in detecting and responding to vehicle limits, exploring optimal driving strategies, and maximizing vehicle potential. These findings provide directions for the future development of autonomous driving systems.

Accelerating Autonomy: Insights from Pro Racers in the Era of Autonomous Racing - An Expert Interview Study

Autonomous Vehicles on the Edge: A Survey on Autonomous Vehicle Racing

Outracing Human Racers with Model-based Autonomous Racing

An Adaptive Human Driver Model for Realistic Race Car Simulations

Towards an Autonomous Test Driver: High-Performance Driver Modeling via Reinforcement Learning

Indy Autonomous Challenge -- Autonomous Race Cars at the Handling Limits

Outracing Human Racers with Model-based Planning and Control for Time-trial Racing

Expertise Affects Drone Racing Performance

DeepRacing: Parameterized Trajectories for Autonomous Racing

AMZ Driverless: The Full Autonomous Racing System

Real-Time Adaptive Velocity Optimization for Autonomous Electric Cars at the Limits of Handling

Race Driver Evaluation at a Driving Simulator using a physical Model and a Machine Learning Approach

BayesRace: Learning to race autonomously using prior experience

A Simulation Benchmark for Autonomous Racing with Large-Scale Human Data

A Review of Full-Sized Autonomous Racing Vehicle Sensor Architecture

An interview study exploring Tesla drivers' behavioural adaptation

Design of an Autonomous Racecar: Perception, State Estimation and System Integration

TUM autonomous motorsport: An autonomous racing software for the Indy Autonomous Challenge

Fast and Modular Autonomy Software for Autonomous Racing Vehicles

An embodied and ecological approach to skill acquisition in racecar driving

A Phenomenological Evaluation of the Psychological Factors That Influence Motorsports Performance