Our project has realized all the expected features and functions listed as below:
(1).
The system can accurately distinguish the crying sound of a baby, and make the right judgement using the machine learning model. Although erroneous judgements are not totally avoided, in which case other noises may be recognized as crying sound, the system can recognize a crying sound correctly as long as there is crying sound. Furthermore, since the amplitude of external noise is greatly attenuated when transmitted into the car, the environment in a car is relatively quiet, which decreases the probability of erroneous judgement of our system.
(2).
The APP can correctly communicate with the server, which means receiving the alert message from the server and sending confirm message to the server can be done without error. The TCP socket technique applied in this communication part guarantees the message packets sent between the server and our APP will not be dropped or corrupted.
(3).
The APP can correctly display the received alert message along with the current in-car temperature, notifying the user how severe the situation is.
(4).
The APP’s “confirm and open fan” function provides the user a counter measure to cooling down the in-car space by turning on the fan. Furthermore, the fan’s operation load is decided by the temperature. If the temperature is lower than 20, the fan runs at lower capacity. If the temperature exceeds 20, then the fan runs at full capacity. After 1 minute the fan is automatically shut down, waiting for the next “open fan” command from the server. Thus this whole process forms a closed loop.
(5).
The system can “cache” the alert message, and displays the message on the cellphone once the user opens the APP. This measure ensures that the user will not miss any alert messages.
(6).
The cellphone receives a SNS alert message along with the alert message to the APP. This measure guarantees the user will be notified even in an area where the network signal is bad and the APP cannot be used. This forms a dual fail-safe.
The project has a promising prospective since almost every family owns at least a car, and they will encounter the similar situation in summer definitely. However, there are no such device in the market which can realize such function. The safety of children is always a big concern and is placed in the top priority to every family. So, our project is designed just at the time.
(1).
The system can accurately distinguish the crying sound of a baby, and make the right judgement using the machine learning model. Although erroneous judgements are not totally avoided, in which case other noises may be recognized as crying sound, the system can recognize a crying sound correctly as long as there is crying sound. Furthermore, since the amplitude of external noise is greatly attenuated when transmitted into the car, the environment in a car is relatively quiet, which decreases the probability of erroneous judgement of our system.
(2).
The APP can correctly communicate with the server, which means receiving the alert message from the server and sending confirm message to the server can be done without error. The TCP socket technique applied in this communication part guarantees the message packets sent between the server and our APP will not be dropped or corrupted.
(3).
The APP can correctly display the received alert message along with the current in-car temperature, notifying the user how severe the situation is.
(4).
The APP’s “confirm and open fan” function provides the user a counter measure to cooling down the in-car space by turning on the fan. Furthermore, the fan’s operation load is decided by the temperature. If the temperature is lower than 20, the fan runs at lower capacity. If the temperature exceeds 20, then the fan runs at full capacity. After 1 minute the fan is automatically shut down, waiting for the next “open fan” command from the server. Thus this whole process forms a closed loop.
(5).
The system can “cache” the alert message, and displays the message on the cellphone once the user opens the APP. This measure ensures that the user will not miss any alert messages.
(6).
The cellphone receives a SNS alert message along with the alert message to the APP. This measure guarantees the user will be notified even in an area where the network signal is bad and the APP cannot be used. This forms a dual fail-safe.
The project has a promising prospective since almost every family owns at least a car, and they will encounter the similar situation in summer definitely. However, there are no such device in the market which can realize such function. The safety of children is always a big concern and is placed in the top priority to every family. So, our project is designed just at the time.