CxBR encapsulates knowledge about appropriate actions and/or procedures, as well
as compatible new situations, into hierarchically organized contexts. Tactical behavior
is action-based at low levels (e.g., keeping the car on the traffic lane), decision-andaction
based at middle levels (stop or go at a traffic light), and decision-based at high
levels (take the freeway or the back roads). In this sense, CxBR is hierarchical and
modular and a sample hierarchy is depicted in Figure 7.5.
The mission context defines the scope of the mission, its goals, the plan, and the
constraints imposed (time, weather, etc). The main context contains functions, rules,
and a list of compatible subsequent main contexts. Identification of a new situation
can now be simplified because only a limited number of all situations are possible
under the currently active context. Subcontexts are abstractions of functions performed
by the main context, which may be too complex for one function or that may
be employed by other main contexts. This encourages re-usability. Subcontexts are
activated by rules in the active main context. They will deactivate themselves upon
completion of their actions.
Decisions are heavily influenced by a sequence of main contexts, each of which,
when active, controls behavior of an autonomous vehicle agent (either real or simulated)
with an expectation for the future. Active main contexts change not only in
response to external events or circumstances but also because of actions taken by the
agent itself. One example of a main context could be driving in city traffic, called
City Driving. Such a context would contain functions to maintain the vehicle on the
road at a speed not to exceed the speed limit, the know-how to handle intersections,
pedestrians, school zones, and so forth. It could call subcontexts to help it deal with
traffic lights, school zones, and emergency vehicles.