I need someone to develop an electronic circuit for a new product I am creating.
It detect people inside the water and the circuit uses a sonar, process the signal and trigger an alarm.
The project needs to include analyses about battery lifetime, accuracy of the output, quality of sound waves etc.
I can give more information and there is also a product in the market using the same technology to detect fishes.
It is really important to realize that the complex part is how to use decide the best use of the sonar. Bellow is the best solution we have now. Please, think about it and give me your opinion or questions:
When the customer buy the product, install it and press the power button for the first time, it will receive the ambient noise (working as a PASSIVE sonar). Then we will have a pattern #1, so we will be able to detect variation on this "natural" ambient.
Then the sonar will send a signal and receive what it gets. This will give the pattern #2 which will be used in a second moment.
So, when a kid or an animal fall down in the pool, we are expecting that we can recognize it (I have to test first and see what it gives to me). And we have to be able to differ between this and a "false alarm" that can be a ball or something else that we don't want to trigger the alarm.
When it gets a positive answer, it will send a signal (now working as a ACTIVE sonar) and compare with the pattern #2. If it also confirms, then it triggers the alarm.
Summarizing... the idea is to work as a passive sonar first (also to preserve the battery) and use the active sonar to confirm it. We just have to see if the time to do this will not compromise the response time.
Another feature we want and we need to check if it is feasible is to use a solar panel and a rechargeable battery. When it is sun, the system uses the energy directly from the solar panel and when it is necessary, it takes from the battery.
Please, tell me what you think about this and if you can see some gap or improve the solution.
WE STILL ANALYZING THE AVAILABLE TECHNOLOGIES. WE TESTED WITH AN ACCELEROMETER, BUT IT DIDNT WORK WELL.