Abstract
Nowadays predictive maintenance is the powerful technical in diverse fields, requiring both state of the art
equipment and highly qualified engineers and technician involved. The role of software used is extremely
important for the success of new technologies. The newest and the most advanced technologies available
appear to have no limits.
Recognizing the growing severity of air, water and ground pollution, caused by industrial and due to welfare
expansion, and considering increasing vehicle population, this study can contribute to restore the better air
quality in the hospitals and at the same time to improve the maintenance management.
The studies of effects of pollution in hospital areas will allow enhance the role of efficient maintenance
management. Sometimes they characterize the problem as how to determine the importance of the observed
levels of pollution to the biological system. One difficulty is a prediction of future consequences of same kinds
of pollutants.
This paper presents a possible route to create ecologically clean methods in asset management, inside hospital
applications.
Rapid simultaneous multi-emission indicators determination, including noise, can to be implemented in hospital
areas, by different detection techniques (atomic absorption spectrometry, mass spectrometry, termography,
noise analysis, ultrasonography, vibration analysis, laser, plasma, particles in lubricating oil). The choice of
appropriate indicators correlated with dysfunctions occurrence is an important step. In addition, the complexity
increases, when the investigation is associated with poor quality of medical air in a hospital location, as
consequence of the urban pollution. The concentration of same gases peaked at rush hours within the day.
In hospital rooms, detection limits and precision must to be defined at low values in way to preserve the correct
services operation.
Our goal tries to join a balanced maintenance administration with the mitigation of indicators of condition with
risk of provoking environmental impacts. In the control absence, such impacts can contribute to the worsening
of hospital healthy conditions.
This strategy offers a useful diagnostic tool to determine the existence of abnormal equipment operation.
Presently, integrated systems used in maintenance management are common tools at the maintenance
departments of industrial organizations, as well as in hospital and hospitality fields, among others. However,
their contribution to the benefit of these organizations is sometimes questionable, as the maintenance
diagnosis is a prerequisite to the success of these systems, to achieve long live and the better operation
conditions. This is the main reason for his introduction in hospital areas, with great probability of much more
successful. At the same time, ought to be recognized that the improvement of the maintenance management
will be achieved through good planning, better organization and the introduction of fault diagnosis systems.
The team that presents this paper is making his research and development around that wide spectrum of a new
vision.
The base of development is an integrated system for hospital terology management, called SMITH (Farinha,
1994)[1], that permits to manage the maintenance in general, including on-condition maintenance, but it is
modular enough to permit to associate new features, namely this last approach, in order to maintain the system
up-to-date and, always possible, to anticipate the future.
In this case, we may use on-line monitoring of state indicators or to generate optimal inspection strategies for
randomly failing system, whose state is hidden. In this way, it is possible to maximize the production cycle and
anticipated some dysfunction. To make this, we use hardware tools that can easily be found in the market,
associated with usual software tools, in order to develop a solution that represents a new approach in this field.
The methodologies are based on Hidden Markov Models, associated the respective algorithms in the SMITH.