The management of public transportation suffers from the difficulty
of predicting population growth and consequently meet its demands
in terms of quality and safety. With regard to safety, several computer
systems have emerged to help prevent and treat transport
problems, such as detour of routes and speeding. However, in the
context of the Manaus city, these systems cannot: (a) detect specific
objects (firearms, knives, among others) to identify assaults; and (b)
identify overcrowding in means of transportation. The inclusion of
these functionalities with the above mentioned ones may allow access
to more information to subsidize decision making and improve
the quality of security offered by the public transportation service.
This article presents a proposal for an intelligent system to monitor
the safety of a bus related to the above mentioned points. In order
to raise the requirements and propose the graphical interface of the
system, the techniques of benchmarking and focus group were used.
The prototype presented allows simulating how the information
would be presented to those responsible for monitoring the bus,
in addition to generating reports for monitoring the history of the
means of transport and to support decision making regarding the
safety of passengers and servers.