Error States

We can utilize the observer framework for recognizing error states and sequences. The idea behind this recognition task is to be able to report on visually incorrect sequences. In particular, if there is a pre-determined observer model of a particular exploration task under observation, then it would be useful to determine if something goes wrong with the exploration actions. The goal of this reporting procedure is to alert the operator or autonomously supply feedback to the exploring robot so that it can correct its actions.

Some examples of errors that might occur while exploring based on a reverse engineered model include:

• occlusions between the observer camera and the part or probe.
• inappropriate approach vector position or orientation (see Figure 5).
• inappropriate probe size (see Figure 6).
• motion too rapid.
• motion too slow (``frozen'' or ``timeout'').

The correct sequences of automata state transitions can be formulated as the set of strings that are acceptable by the observer automaton. This set of strings represents precisely the language describing all possible visual task evolution steps.