Heat pumps have been toted as a technology will the potential to meet climate targets and support buildings with undergoing the energy transition – but they are not without their faults. This is where embedded AI fits in.
Signs of wear on the electric motor that drives the compressor and faults in the form of too-low or too-high temperature differences in the evaporator and condenser can all cause the heat pump to work inefficiently and produce too little power. For example, the temperature of the coolant could still be too high when it leaves the condenser, so the difference in the energy source is too small when it enters the evaporator. As a result, the coolant can absorb less energy from the environment.
“In addition to a high-pressure and low-pressure fault, other faults can occur in the refrigerant circuit that can cause the heat pump to fail. For example, insufficient refrigerant leads to a drop in the suction pressure in the evaporator,” said Viacheslav Gromov, Managing Director of AI solution provider AITAD. “A defective expansion valve leads to a sub-optimal refrigerant flow between the high-pressure and low-pressure sides, resulting in temperature differences that are too low or too high, meaning that energy can no longer be recovered efficiently. Other typical sources of faults include defective evaporator fans, defective defrosting, dirty registers, closed shut-off valves, and blocked filter dryer.”
AI can be an invaluable tool when it comes to monitoring potential weak points in the heat pump. Critical conditions can be detected and eliminated at an early stage before they occur with the help of intelligent sensor technology. Maintenance of the heating system can therefore be planned conveniently and efficiently without unplanned downtime; predictive maintenance.
On the technology side, the use of intelligent sensors (embedded AI system components), where the AI is housed directly on the sensor board and the heat pump can be monitored without an Internet or Cloud connection, is a good option. Embedded AI also has the great advantage that it processes much larger amounts of data, up to several terabytes per day, which is not possible with conventional cloud or server solutions, as such large amounts of data are almost impossible to transfer.
Use cases for predictive maintenance of the heat pump include:
- Predictive maintenance with ultrasonic/vibration embedded AI sensor on the compressor
- Optimisation/efficiency increase of the compressor through status and anomaly detection by sound/ultrasonic embedded AI sensor technology on the compressor
- Optimisation/efficiency increase of the condenser/evaporator process through status and anomaly detection (heat distribution) by IR grid embedded AI sensor in front of condenser/evaporator
- Optimisation/efficiency increase of the expansion valve through status and anomaly detection (heat distribution) by sound/ultrasonic embedded AI sensors on the valve
- Predictive maintenance with ultrasonic/current/vibration embedded AI sensor on the cooling fan
- Voice and gesture control as well as person recognition optical/radar/lidar embedded AI sensor and indirectly e.g. via VOC embedded AI sensor – the number of people, window opening, etc.
Embedded AI can be used to intelligently control the entire heating system, for instance by regulating the temperature in the room depending on the number of people present, or by recognising whether the windows are open so the system can autonomously adjust itself to the situation.
Embedded AI is the latest megatrend within AI, relating tothe merging of sensors and AI on one board. In doing so, the sensor data is evaluated on site and in real time. Instead of the data, evaluation results are passed on.
The main areas of application are predictive maintenance, user interaction and functional innovations. Due to the limited resources of embedded AI – low memory and energy requirements – the latter can even be covered by energy harvesting in a heat pump – not only is the service life extended, but energy generation is also optimised.
“The heat pump is a promising tool for climate protection and sustainable energy supply. Artificial intelligence can be used to increase the efficiency and service life of the heat pump reliably and with customer benefits. This environmentally friendly technology becomes even more interesting as it gives the heat pump ‘built-in investment protection’. In this way, embedded AI contributes to sustainability and thus ultimately to the success of the energy transition,” said Gromov.
There’s plenty of other editorial on our sister site, Electronic Specifier! Or you can always join in the conversation by commenting below or visiting our LinkedIn page.
