Results of the summarizing process for the arXiv paper: 2104.06262v4

In the field of autonomous vehicles, game engines are becoming popular as simulation platforms for developing vehicle control systems and test environments. However, a crucial requirement for simulation-based development and verification is determinism. A deterministic process always produces the same output given the same initial conditions and event history, which allows for repeatable tests and reliable debugging. Unfortunately, game engines are often non-deterministic, meaning they do not consistently produce the same results under identical conditions. This paper aims to review and identify the potential causes of non-deterministic behaviors in game engines. It specifically focuses on CARLA, an open-source autonomous driving simulation environment powered by Unreal Engine, to highlight its inherent shortcomings in providing precise experimental results. The study explores different configurations and utilization of both software and hardware to determine an operational domain where simulation precision is sufficiently low. In other words, they aim to find a range of variance between repeated executions that becomes negligible for development and testing purposes. To address this issue more broadly, the paper proposes a general method that can be used to identify permissible variances in game engine simulations for any given system configuration. This method could potentially help researchers and developers establish reliable simulation environments with consistent results. Overall, this research highlights the challenges posed by non-determinism in game engines used for autonomous vehicle simulations. By identifying potential causes and proposing a method to mitigate these issues, it contributes to improving the trustworthiness and reliability of simulation-based development and verification processes in this field.

• - Game engines are popular for developing and testing autonomous vehicle control systems
• - Determinism is crucial for simulation-based development and verification
• - Game engines are often non-deterministic, meaning they do not consistently produce the same results under identical conditions
• - The paper focuses on CARLA, an open-source autonomous driving simulation environment powered by Unreal Engine, to highlight its shortcomings in providing precise experimental results
• - The study explores different configurations and utilization of software and hardware to determine an operational domain with sufficiently low simulation precision variance
• - A general method is proposed to identify permissible variances in game engine simulations for any given system configuration
• - This research contributes to improving the trustworthiness and reliability of simulation-based development and verification processes in the field of autonomous vehicles.

Game engines are used to make and test self-driving cars. Determinism means that things happen the same way every time. Game engines sometimes don't do things the same way every time. The paper talks about CARLA, a computer program for testing self-driving cars, and how it has some problems with getting exact results. The study looks at different ways to use software and hardware to make sure the tests are accurate. They also have a way to figure out how much difference is okay in the tests. This research helps make sure self-driving cars are safe and reliable." Definitions- Game engines: Computer programs used to create and test games or simulations. - Autonomous vehicle control systems: Systems that allow vehicles to drive themselves without human input. - Determinism: The idea that events always happen in the same way under identical conditions. - Simulation-based development and verification: Using computer simulations to develop and test new technologies or systems. - Non-deterministic: Not consistently producing the same results under identical conditions. - Open-source: Software that is freely available for anyone to use, modify, and distribute. - Unreal Engine: A popular game engine used for creating realistic graphics and physics in video games. - Experimental results: The outcomes or findings from a scientific experiment or study. - Precision variance: Differences or variations in the accuracy or exactness of measurements or calculations. - Trustworthiness: Being reliable, dependable, and deserving of trust.

Simulation-Based Development and Verification of Autonomous Vehicles: The Challenges of Non-Determinism in Game Engines

The development and verification of autonomous vehicles is a complex process that requires reliable simulation environments. Unfortunately, game engines are often non-deterministic, meaning they do not consistently produce the same results under identical conditions. This paper aims to review and identify the potential causes of non-determinism in game engines, specifically focusing on CARLA, an open-source autonomous driving simulation environment powered by Unreal Engine. It also proposes a general method that can be used to identify permissible variances in game engine simulations for any given system configuration.

What is Determinism?

Before exploring the challenges posed by non-determinism in game engines used for autonomous vehicle simulations, it’s important to understand what determinism means. A deterministic process always produces the same output given the same initial conditions and event history, which allows for repeatable tests and reliable debugging. In other words, determinism ensures consistent results when running experiments or tests multiple times with identical parameters.

Non-Deterministic Behaviors in Game Engines

Game engines are becoming popular as simulation platforms for developing vehicle control systems and test environments due to their flexibility and ease of use; however, they often lack determinism due to their inherent design features such as random number generation algorithms or unpredictable physics calculations. As a result, these features can lead to discrepancies between repeated executions even when all parameters remain unchanged. To address this issue more broadly, this research focuses on CARLA – an open source autonomous driving simulator powered by Unreal Engine – to highlight its shortcomings in providing precise experimental results. Specifically, it explores different configurations and utilization of both software (e.g., operating system) and hardware (e.g., GPU) to determine an operational domain where simulation precision is sufficiently low so that any variance between repeated executions becomes negligible for development and testing purposes.

Proposed Methodology

The paper proposes a general method that can be used to identify permissible variances in game engine simulations for any given system configuration regardless of platform or application type (i.e., autonomous vehicles). This method involves running multiple trials with varying input values while monitoring performance metrics such as frame rate or latency over time; if there is no significant difference between trial runs then it can be assumed that the variance is within acceptable limits for development purposes . Additionally , researchers should consider factors such as hardware specifications , operating system version , compiler optimization levels , etc . before conducting experiments .

Conclusion

Overall , this research highlights the challenges posed by non - determinism in game engines used for autonomous vehicle simulations . By identifying potential causes and proposing a method to mitigate these issues , it contributes towards improving trustworthiness reliability of simulation - based development verification processes .