Belief space planning improves error prediction under uncertain observations
Belief-Space Planning with Planner-Conditioned Estimator Error under Intermittent Observations
Robotics
Summary
Planning safe actions requires understanding how reliable your knowledge about the situation is. The authors found that traditional estimators can miss biases in their errors when they have less or different information than the planner. They created a planning method that predicts and accounts for these biased errors when observations come unpredictably. Their tests with a drone landing on a ship showed more accurate error forecasts and safer landings compared to older methods.
What this means in practice
- •For autonomous vehicle developers: Enable safer and more reliable vehicle path planning by incorporating realistic error predictions during periods of infrequent sensor updates.
- •For drone operation teams: Improve precision in drone landing sequences on moving platforms by better forecasting estimator errors under intermittent observations.
Authors
Peter M. Donley, Joseph L. Moore
Abstract
Belief-space planners with separately designed or off-the-shelf estimators may have access to state-relevant information the estimator does not observe. Consequently, even an estimator that minimizes mean-squared error under its own information can have a nonzero error mean when conditioned on planner information. When future corrections are intermittent and stochastic, differences between accepted and rejected error means introduce additional uncertainty terms to correction events that zero-mean assumptions ignore. In this paper, we propose a belief-space planning approach that predicts, propagates, and penalizes planner-conditioned estimator error along evaluated trajectories. A planner-conditioned estimator-error model combines affine error dynamics with a moment recursion that preserves stochastically-induced correction uncertainty. We describe a method by which to predict estimator error dynamics within specified operating regimes and incorporate predicted error moments into a task-weighted quadratic risk objective. We evaluate the approach in simulation for a tilt-rotor VTOL landing on a ship deck using receding-horizon planning. We find that conditioning on planner information reduced error-prediction loss for a command-blind EKF by 13.4% relative to a planner-ignorant model, with strongly regime and horizon-dependent forecasting abilities. In a small set of 16 paired closed-loop trials, the planner lowered the median terminal task gauge from 1.60 to 0.99, and predicted estimator error beyond 2s within a factor of 1.3, versus a 5.7-fold underprediction by covariance-only planning.