Training Champion-level Race Car Drivers Using Deep Reinforcement Learning

Many potential applications of artificial intelligence involve making real-time decisions in physical systems. Automobile racing represents an extreme case of real-time decision making in close proximity to other highly-skilled drivers while near the limits of vehicular control. Racing simulations, such as the PlayStation game Gran Turismo, faithfully reproduce the nonlinear control challenges of real race cars while also encapsulating the complex multi-agent interactions. We attack, and solve for the first time, the simulated racing challenge using model-free deep reinforcement learning. We introduce a novel reinforcement learning algorithm and enhance the learning process with mixed scenario training to encourage the agent to incorporate racing tactics into an integrated control policy. In addition, we construct a reward function that enables the agent to adhere to the sport's under-specified racing etiquette rules. We demonstrate the capabilities of our agent, GT Sophy, by winning two of three races against four of the world's best Gran Turismo drivers and being competitive in the overall team score. By showing that these techniques can be successfully used to train championship-level race car drivers, we open up the possibility of their use in other complex dynamical systems and real-world applications.

Assessment of Dietary Intake of Long-Distance Race Car Drivers—A Pilot Study

Long-distance race car drivers are classified as athletes. The sport is physically and mentally demanding, requiring long hours of practice. Therefore, optimal dietary intake is essential for health and performance of the athlete. The aim of the study was to evaluate dietary intake and to compare the data with dietary recommendations for athletes and for the general adult population according to the German Nutrition Society (DGE). A 24-h dietary recall during a competition preparation phase was obtained from 16 male race car drivers (28.3 ± 6.1 years, body mass index (BMI) of 22.9 ± 2.3 kg/m2). The mean intake of energy, nutrients, water and alcohol was recorded. The mean energy, vitamin B2, vitamin E, folate, fiber, calcium, water and alcohol intake were 2124 ± 814 kcal/day, 1.3 ± 0.5 mg/day, 12.5 ± 9.5 mg/day, 231.0 ± 90.9 ug/day, 21.4 ± 9.4 g/day, 1104 ± 764 mg/day, 3309 ± 1522 mL/day and 0.8 ± 2.5 mL/day respectively. Our study indicated that many of the nutrients studied, including energy and carbohydrate, were below the recommended dietary intake for both athletes and the DGE.

Redox Changes in Amateur Race Car Drivers Before and After Racing

Despite the unique opportunity race car driving provides to study exercise in extreme conditions, the sport of racing is under-represented. A better understanding of how racing changes physiological measures combined with driver demographics may help reduce driver risks and expand the field of driver science. This study charted the changes in heart rate, body temperature, blood pressure, static oxidation reduction potential (sORP), and antioxidant capacity in drivers before and after racing (n=23). The interaction between racing and driver characteristics on physiological variables were evaluated. Heart rate, body temperature, and sORP were elevated after racing (P<0.05). Age, cockpit temperature, experience, and speed did not correlate with physiological or oxidative measures (P>0.05). Elevated post-race sORP values were associated with higher pre-race systolic blood pressure and lower antioxidant capacity (P<0.05). We conclude that racing alters the redox response in drivers and that drivers’ pre-race systolic blood pressure and antioxidant capacity can further alter it. A better understanding of the physical and oxidative changes which result from racing may help minimize the unique risks