As I mentioned in a previous message, I contacted DJI support and it seems that it is actually being fruitful (thanks DJI). One of the things I asked them about was about not being able to detect IR shots and this is the information I got back:
For detecting and receiving infrared attacks, you can refer to the "armor_hit_detection_top_right(MSG)" and the "armor_hit_detection_top_left(MSG) " API, https://www.dji.com/robomaster-s1/programming-guide.

Assuming I understand this correctly, what they are stating is that IR hits are registered only in the gimbal (that is unexpected at least to me).

So, can any of you guys try this? I am at work without access to my S1s now and without power at home (PG&E shut down power to a considerable chunk of the bay area).

If you are able to try it and see if it really works, please report back!

Also, I opened a formal bug about being unable to fire shots in autonomous mode. They are looking into it so I would expect a fix in the next release.

In retrospect, this should have been obvious after Duane Degn's tear-down of the gimbal and chassi hit detectors. Only the gimbal has IR receivers.

Here is a simple program to check if it works. 2 S1s would be needed to test it (only one needs to run the program, of course).

def start():
  while True:
    time.sleep(60)
   
def armor_hit_detection_top_right(msg):
  report_hit("right side")

def armor_hit_detection_top_left(msg):
  report_hit("left_side")

def report_hit(location):
  print(f'Hit detected at {location}.')

In theory, if an IR hit is detected, the console will show the side that detected it. This should also work for gel bead hits in the gimbal sides.

BGA Posted at 10-10 08:49
In retrospect, this should have been obvious after Duane Degn's tear-down of the gimbal and chassi hit detectors. Only the gimbal has IR receivers.

"Only the gimbal has IR receivers."

For anyone who hasn't seen these photos, here's the IR receivers in the gimbal.



I'm guessing the IR LEDs are used to identify which robot was hit. I'm not at all sure about this.

Has anyone checked to see if the IR hits register well outside in direct sunlight? I know the sun can cause all sorts of problems because the IR from the sun swamps the receivers.

Here''s a photo of the inside of the hit detectors.

BGA Posted at 10-10 10:17
Here is a simple program to check if it works. 2 S1s would be needed to test it (only one needs to run the program, of course).

def start():

Ok, it does not work. I reported it back to DJI.

Ok, that makes sense.  The IR hit detectors are arranged in an out, forward, back configuration on two sides.  This makes sense given the wide and large "blast pattern" of the IR Blaster gun.  No need for any more IR hit detectors than that.  I'll try to confirm the no-worky result on that code snippet.

Are we able to have a support on the IR signal to use to simulate a shooter?

GG38 Posted at 11-3 23:12
Are we able to have a support on the IR signal to use to simulate a shooter?

Not sure I understand your question but, in any case, you can fire the IR blaster programmatically  in Solo or Battle modes by running the program as a skill and having IR set as the firing mode. But even with this, you can not programmatically detect IR hits. There is simply no support in the current API.

Ok, clear enough

Ok, so I was wondering about the robots detecting infrared signals. I have 2 robots that are used by me and my dad and my robot detect infrared hits on every panel. However, his bot doesn't detect hits anywhere. I have tried hitting the gimbal and all the panels and we even entered a free for all mode to test it out. I thought it was just a bug, so I booted up the bots and connected again, but this time neither of the bots could receive any hits at all. I was wondering if anyone could possibly identify the issue and help me solve it. Thanks.

ThatDoodNate Posted at 4-15 11:52
Ok, so I was wondering about the robots detecting infrared signals. I have 2 robots that are used by me and my dad and my robot detect infrared hits on every panel. However, his bot doesn't detect hits anywhere. I have tried hitting the gimbal and all the panels and we even entered a free for all mode to test it out. I thought it was just a bug, so I booted up the bots and connected again, but this time neither of the bots could receive any hits at all. I was wondering if anyone could possibly identify the issue and help me solve it. Thanks.

Again, only the turrets have ir receivers. The armor does not have those. Note that for an S1 to detect IR hits from another S1, it has to also have IR set as the firing mode.

Also, since the latest firmware, detecting IR shots can be done while running a program.

Actually, I tested what you said about the gimbal and that isn't true. In my model, the armor can detect the IR signals. And In battle mode, I made sure it was set to infrared mode. I tested once again after posting my comment and again, the robot that I was using can detect the signals on the armor and gimbal, but the other one can't detect them at all.

ThatDoodNate Posted at 4-15 13:02
Actually, I tested what you said about the gimbal and that isn't true. In my model, the armor can detect the IR signals. And In battle mode, I made sure it was set to infrared mode. I tested once again after posting my comment and again, the robot that I was using can detect the signals on the armor and gimbal, but the other one can't detect them at all.

I am not sure how you tested it, but I am sure you are not testing what you think you are testing. The only IR receivers in the S1 are in the gimbal.

If you mean that in the app the feedback looks like it indicates that the hit was in an specific armor, that is just a software truck. As there are 3 receivers per side, you can measure the timing in all sensors that detected the hit to figure out the direction it came from and, with this information, say that armor X was hit.

Can it really detect a timing difference between the IR receivers?  I would think since IR is travelling at the speed of light, it would be hard to detect different timings with distances from emitter to receivers only being a few meters max and the distance between receivers only a few CM apart.  Detected intensity would more likely be the criteria, where most of the time only one detector out of the 6 would pick up a beam.

One way for ThatDoodNate to validate what's detecting the IR is to cover the turret's detectors with black tape.  I bet then there would be no detection.

DanMan32 Posted at 6-5 12:38
Can it really detect a timing difference between the IR receivers?  I would think since IR is travelling at the speed of light, it would be hard to detect different timings with distances from emitter to receivers only being a few meters max and the distance between receivers only a few CM apart.  Detected intensity would more likely be the criteria, where most of the time only one detector out of the 6 would pick up a beam.

One way for ThatDoodNate to validate what's detecting the IR is to cover the turret's detectors with black tape.  I bet then there would be no detection.

Well, it most likely would require more expensive sensors but it is not technically impossible. In any case, the S1 most likely just uses the angle of incidence and  which sensor was triggered to figure out direction.

Anyway, as I mentioned, we know for a fact that the only IR sensors in the S1 are in the gimbal so, yes, your tests would show exactly this (no hits would be detected).