ArduinoTroubleshooting

C1001 - SEN0623 - mmwave Human Presence dectector

userHead BScott 2024-08-30 02:51:08 369 Views3 Replies

I have the C1001 board and even after getting the examples working and reading the Wiki, I still have some questions!

 

I'd like to know what radar sensor this product uses, so I can perhaps dig deeper into how to communicate with it.  The included library has a few bugs and at least one misspelling ("gitHeartRate(void)"….?) which suggests that it is a revision of another library, perhaps from the original radar manufacturer.

 

I did not really expect the chip to accurately read respiration and pulse, much less estimate sleep “quality”, but I did hope it could tell the difference between a human lying down and standing up - so far, I can't get it to do that.  I've only tried the sleep mode but based on my experience I do not have high hopes for the fall-detection mode either!

 

A few specific questions:

- there's a red LED which lights up when a human is present - whether or not it is “set up” in the code.  This works great.  But hu.getLEDLightState(hu.eHPLed) always returns a 1- is it even POSSIBLE for comprehensiveState.presence and hu.smSleepData(hu.eInOrNotInBed) to report different values?  I expect the first to give me a 1 if something roughly human-sized is within range, and the second to give me a 1 only when that object is roughly horizontal…- I understand that the chip probably waits a while after detecting me (the LED comes on) and reporting the presence of a human (via comprehensiveState.presence) but how long does it wait before trying to guess whether I'm asleep?  How does it decide to begin to (claim to) measure pulse/respiration?- comprehensiveState.averageRespiration and comprehensiveState.averageHeartbeat both return numbers when measuring your vitals, but when no one is around or before they begin doing that, they return different values - averageHeartbeat is a 0, which makes sense, but averageRespiration can't be converted to a number??- once it begins measuring vitals, it does not stop reporting them even after no one is in the room…

 

I've purchased and experimented with a number of radar sensors and so far this one is no better than the others, although it cost 5x more.

 
2024-09-09 16:58:36

I'm also having trouble getting useful data from this sensor.

 

Using the example from https://github.com/DFRobot/DFRobot_HumanDetection/blob/master/examples/basics/basics.ino

 

In a completely empty room, I get:

 

“Existing information:No one is present”

 

If I induce a (non-human) slight movement in the room, I get

 

"Existing information:Someone is present"
 

for 60 seconds. Well that isn't useful… The whole point of this expensive sensor is to detect humans, not objects. This is no better than a PIR sensor that costs a small fraction of the price.

 

 

The respiration and heart rate values seem to be basically completely made-up.

 

I put the sensor on a table 1m from me, pointed right at my chest. Maybe eventually after a few minutes, it will start to report respiration and heart rates…  Or maybe not. When it does, they're only vaguely accurate: +/- 20 beats from my actual measured heart rate, for example. I don't expect perfect accuracy, +/- 20 beats is actually okay, but the fact that it takes so long to notice me in the first place is a big problem.

 

An even bigger problem: after it starts eventually reporting respiration and heart rates, if I then leave the room, it will continue to report completely fictional changing respiration and heart rates for another 60 seconds! It's not just that it's reporting the last observation, it's new numbers every reading, sometimes climbing to nearly double what my actual heart rate was when I left.

 

 

My overall impression of this sensor is very poor. It costs a lot ($35 US), and in the end provides no useable information beyond what a PIR sensor ($1 US) could tell me. I deeply regret my purchase.

 

 

userHeadPic RoboCurious
RoboCurious wrote:

Even more laughable: if I set up something (not human) in the room to move periodically (every 20-30 seconds), it will register “Body movement parameters” that keep the “Someone is present” message alive, and then also … CONTINUE TO INVENT COMPLETELY FAKE HEART AND RESPIRATION RATE DATA.

 

This sensor is a complete fraud.

 

 

2024-09-09 17:11:16
1 Replies
2024-08-30 03:04:19

OK can I edit this to have better line breaks…?

userHeadPic BScott