Robust Forecasting for Robotic Control: A Game-Theoretic Approach
This addresses the need for reliable predictions in robotics, such as for self-driving cars, by mitigating issues like noise and outliers in historical data, though it is incremental as it builds on existing forecasting methods with a novel adversarial twist.
The paper tackles the problem of generating robust forecasts for robotic control by proposing a game-theoretic framework where an adversary perturbs historical time series to increase control costs, resulting in a 30.14% improvement on out-of-distribution real-world lane change data compared to baselines.
Modern robots require accurate forecasts to make optimal decisions in the real world. For example, self-driving cars need an accurate forecast of other agents' future actions to plan safe trajectories. Current methods rely heavily on historical time series to accurately predict the future. However, relying entirely on the observed history is problematic since it could be corrupted by noise, have outliers, or not completely represent all possible outcomes. To solve this problem, we propose a novel framework for generating robust forecasts for robotic control. In order to model real-world factors affecting future forecasts, we introduce the notion of an adversary, which perturbs observed historical time series to increase a robot's ultimate control cost. Specifically, we model this interaction as a zero-sum two-player game between a robot's forecaster and this hypothetical adversary. We show that our proposed game may be solved to a local Nash equilibrium using gradient-based optimization techniques. Furthermore, we show that a forecaster trained with our method performs 30.14% better on out-of-distribution real-world lane change data than baselines.