Space free-flyers like the Astrobee robots currently operating aboard the International Space Station must operate with inherent system uncertainties. Parametric uncertainties like mass and moment of inertia are especially important to quantify in these safety-critical space systems and can change in scenarios such as on-orbit cargo movement, where unknown grappled payloads significantly change the system dynamics. Cautiously learning these uncertainties en route can potentially avoid time- and fuel-consuming pure system identification maneuvers. Recognizing this, this work proposes RATTLE, an online information-aware motion planning algorithm that explicitly weights parametric model-learning coupled with real-time replanning capability that can take advantage of improved system models. The method consists of a two-tiered (global and local) planner, a low-level model predictive controller, and an online parameter estimator that produces estimates of the robot’s inertial properties for more informed control and replanning on-the-fly; all levels of the planning and control feature online update-able models. Simulation results of RAT-TLE for the Astrobee free-flyer grappling an uncertain payload are presented alongside results of a hardware demonstration showcasing the ability to explicitly encourage model parametric learning while achieving otherwise useful motion.

The performance of model-based motion control for free-flying robots relies on accurate estimation of their parameters. In this work, a method of rigid body inertial parameter estimation which relies on a variational approach is presented. Instead of discretizing the continuous equations of motion, discrete dual quaternion equations based on variational mechanics are used to formulate a linear parameter estimation problem. This method depends only on the pose of the rigid body obtained from standard localization algorithms. Recursive semi-definite programming is used to estimate the inertial parameters (mass, rotational inertia and center of mass offset) online. Linear Matrix Inequality constraints based on the pseudo-inertia matrix ensure that the estimates obtained are fully physically consistent. Simulation results demonstrate that this method is robust to disturbances and the produced estimates are at least one order of magnitude more accurate when compared to discretization using finite differences.

The mobile manipulation of arbitrary objects in micro-gravity by a free-flying robot is a challenging problem. One particular problem is the control of the robot after grasping an object of unknown inertial properties. This paper addresses the problem of determining, while the robot is grasping an object, the inertial properties of the resulting robot-object body. This is done by first optimizing an exciting trajectory, and then estimating the inertial parameters from the executed trajectory. These trajectories are tracked using a Nonlinear Model Predictive Control (NMPC) controller with actuation constraints. Both planned and executed trajectories are represented using Fourier series. The approach is evaluated in simulation using the Space CoBot free-flying robot.