Hidden Markov Models as a Process Monitor in Robotic Assembly

Abstract

A process monitor for robotic assembly based on hidden Markov models (HMMs) is presented. The HMM process monitor is based on the dynamic force/torque signals arising from interaction between the workpiece and the environment. The HMMs represent a stochastic, knowledge-based system in which the models are trained off-line with the Baum-Welch reestimation algorithm. The assembly task is modeled as a discrete event dynamic system in which a discrete event is defined as a change in contact state between the workpiece and the environment. Our method (1) allows for dynamic motions of the workpiece, (2) accounts for sensor noise andfriction, and (3) exploits the fact that the amount of force information is large when there is a sudden change of discrete state in robotic assembly. After the HMMs have been trained, the authors use them on-line in a 2D experimental setup to recognize discrete events as they occur. Successful event recognition with an accuracy as high as 97% was achieved in 0.5-0.6 s with a training set size of only 20 examples for each discrete event.

Bayesian time-series models for continuous fault detection and recognition in industrial robotic tasks

“…In robotics, HMMs and their extensions have been successfully applied to encode human motion primitives and transfer them to a humanoid robot [10], to programming by demonstration [11], and to model contact events in assembly tasks [1] . The HMM can be seen as a stochastic finite state automaton, where each state emits an observation.…”

“…The first category comprises approaches that focus on modelling events that discretise the task execution in sub-tasks relevant for control purposes. In [1], for example, Hidden Markov Models (HMMs) are used to model contact events between an object manipulated by a robot and the environment. By analysing the force/torque spectrogram, the type of contact can be identified among a discrete set of edge-surface configuration possibilities.…”

Robot introspection with Bayesian nonparametric vector autoregressive hidden Markov models

“…Namely the Hidden Markov model or variants thereof. In [11], Hovland and McCarragher pioneered the use of HMMs to model contact events by observing wrench signatures. The contact state was identified among a set of discrete edge-surface configurations and provided a probability over a sequence of contacts.…”

Polyhedral contact formation identification for autonomous compliant motion: exact nonlinear bayesian filtering

“…In the past, random walks have been sporadically used and only in particular situations; for example, to escape local minima in potential field methods [9]. Instead, our approach relies entirely on this concept.…”