Page 7 - InnoRenew CoE International Conference 2023
P. 7
Fault detection and diagnostics
Dušan Janjić1,*
1 ROBOTINA D.O.O., OIC Hrpelje 38, 6240 Kozina, Slovenia, dusan.janjic@robotina.com
* Corresponding author
Building automation can significantly contribute to a building’s energy efficiency, comfort, and security, not only directly
through automated procedures and schedules, but also through visualization and especially data analysis of data collected
during building operation. When we talk about visualization, we usually have in mind nice touch screens to operate various
devices in different spaces, but with the right application/automation behind visualization can alert us to unusual use of
devices, consumption etc. Data analysis is even more interesting, and it comes on different levels.
Fault detection and diagnostics is a solution that fits into the data analysis. It not only follows the operation of devices, but
also measures the impact of the device’s action. If this impact is comparable with the baseline, then the system assumes
that the device is working properly. If the impact is not what is expected, then a fault is reported. Fault is not necessarily
an error and, in this case, can be just a deviation of expected outcome from the device, but the efficiency of the device is
reduced. The consequence of reduced efficiency is higher energy usage, less environmental friendliness, and higher costs.
The reasons for the fault can be different and statistics show that 45% of such faults in buildings can be resolved in less
than a minute. As an example, we can give an obstruction (p.ex. plastic bag) in the ventilation system. Automatics of the
system will increase the RPMs of the system to keep the desired air flow, but this will also increase power consumption.
Another example is when the heating in a certain room is turned on. The expected outcome is that the room will heat 1
degree every 2 minutes. If the room temperature after 5 minutes is not raised 2.5 degrees fault is triggered.
Predictive maintenance is also part of this solution. If we take our example of ventilation. This time we notice higher power
consumption to keep the same RPMs. Worn bearings could be the reason. We can plan for repairs before the device fails
completely. In both cases we will have a higher cost of energy, repairs, and worse environmental impact.
Keywords: building automation, fault detection and diagnostics, predictive maintenance
13–14 SEPTEMBER 2023 I IZOLA, SLOVENIA 7
Dušan Janjić1,*
1 ROBOTINA D.O.O., OIC Hrpelje 38, 6240 Kozina, Slovenia, dusan.janjic@robotina.com
* Corresponding author
Building automation can significantly contribute to a building’s energy efficiency, comfort, and security, not only directly
through automated procedures and schedules, but also through visualization and especially data analysis of data collected
during building operation. When we talk about visualization, we usually have in mind nice touch screens to operate various
devices in different spaces, but with the right application/automation behind visualization can alert us to unusual use of
devices, consumption etc. Data analysis is even more interesting, and it comes on different levels.
Fault detection and diagnostics is a solution that fits into the data analysis. It not only follows the operation of devices, but
also measures the impact of the device’s action. If this impact is comparable with the baseline, then the system assumes
that the device is working properly. If the impact is not what is expected, then a fault is reported. Fault is not necessarily
an error and, in this case, can be just a deviation of expected outcome from the device, but the efficiency of the device is
reduced. The consequence of reduced efficiency is higher energy usage, less environmental friendliness, and higher costs.
The reasons for the fault can be different and statistics show that 45% of such faults in buildings can be resolved in less
than a minute. As an example, we can give an obstruction (p.ex. plastic bag) in the ventilation system. Automatics of the
system will increase the RPMs of the system to keep the desired air flow, but this will also increase power consumption.
Another example is when the heating in a certain room is turned on. The expected outcome is that the room will heat 1
degree every 2 minutes. If the room temperature after 5 minutes is not raised 2.5 degrees fault is triggered.
Predictive maintenance is also part of this solution. If we take our example of ventilation. This time we notice higher power
consumption to keep the same RPMs. Worn bearings could be the reason. We can plan for repairs before the device fails
completely. In both cases we will have a higher cost of energy, repairs, and worse environmental impact.
Keywords: building automation, fault detection and diagnostics, predictive maintenance
13–14 SEPTEMBER 2023 I IZOLA, SLOVENIA 7
