Bayesian Hybrid Model-State Estimation applied to Simultaneous Contact Formation Detection and Geometrical Parameter Estimation

This paper describes a Bayesian approach to model selection and state estimation for sensor-based robot tasks. The approach is illustrated with an example from autonomous compliant motion: simultaneous contact formation recognition and estimation of geometrical parameters. Previous research in this area mostly tries to solve one of the two subproblems, or treats the Contact Formation recognition problem separately, avoiding interaction between the Contact Formation detection and the geometrical parameter estimation problems. This limits the application area to task execution under small uncertainties. The problem shows similarities with the well known problems of data association in SLAM and model selection in global localisation. The paper discusses an experiment in which the performances of two well known Bayesian algorithms are compared with respect to this problem: Kalman Filter variants and Particle Filter solutions. This research allows the robot to handle large uncertainties during the execution of its sensor-based task through the estimation of a hybrid joint density of both unknown model and state variables.