Our team installed vision sensors on the robot to identify the Mobile Goals.
However, there were times when we couldn’t tell if the vision sensor was working.
For example, I let the vision sensor light up different colors of lights when Identifying different colored Mobile Goal. But in fact, the lights are not on when running to that step, so the robor can’t ldentify the Mobile Goal.
I checked the log of the master controller and found that the vision sensor is often disconnected, but the light on the vision sensor is always on, I can’t find the reason for the disconnection. I replaced the vision sensor, cable and port with a new one After that, I still have the same problem. When I checked the seven robots we have here and found that they all have the same problem, I was puzzled.
I can vouch that my program is correct because it has been run successfully when the vision sensor is normal.
So, I don’t know if you have the same problem, I don’t know if it’s a bug in the vision sensor or if I’m doing something wrong, please let me know, thanks.
（I use VEXcode Pro V5 to program）
I’ve been using the vision sensor since last season and I think it recognizes it very well, but it just doesn’t work from time to time.
And my one minute auto is going to be pretty bad if I don’t use the vision sensor at all
My team gets those notifications sometimes. One thing our team did was make the brain screen flash a colour depending on what it saw. when it does engage, it flashes red. We did this so we can tell weather the vision or code is malfunctioning. hope this helps. : )
im sorry to report but the vision sensor is more pain than its worth. ive tryed tomake it work but its too inconsistent to be used. it doesnt sense a mobile goal but instead senses the entire field as one huge mobile goal. its best to not stress over it and find alterative methods to make your robot.
I think one issue is that the lighting changes from venue to venue, making it a complete pain to use. I would just use tracking wheels + inertial sensor, it’s far better really. If you really want to, you could also use GPS for skills auton.
While this may not fix everything, just check your brain screen for disconnects before you start the round. its what we do and we haven’t had any issues midround
The vision sensor can be a bit finicky at times, but don’t get discouraged! I absolutely love using them, my team’s robot this year has two of them, they’re useful for autonomous skills, game autons, and even as a driver assist during matches. I’d suggest you first ensure that your signatures are tuned for the current lighting conditions your bot is operating in and that you’re getting accurate readings for all the different goal colors, this is usually the primary reason for inaccuracies, especially if the vision sensor light is on. If you’re still having troubles, try swapping ports on the brain, sometimes ports just act a bit strange. Let me know how it goes, if you encounter any other issues I’m happy to help you look further into it.
Best of luck with getting it working, as said, vision can be strange at times, but once you have it down it’s more than worth the effort.
These are symptoms of an improperly tuned vision sensor; once the tuning is down to the specific lighting where your bot is running, vision is very accurate and reliable in detecting goals of all different colors. I would especially recommend messing with the brightness setting, as that significantly affects performance. Change the sensitivity for the colors as well, a combination of these two is what really gets the performance up to spec.
Mine one minute auto uses vision sensors to control yaw, so I have to keep using it. My problem is that I can’t tell if the vision sensor is in a normal state and I don’t know why it’s like this, is this a bug in the system or where am I doing something wrong?
My vision sensor doesn’t work in many cases and everything works fine.
And the vision sensor doesn’t recognize and light up the color I programmed after the green light is on at power on.
I don’t know why it turned out like this, I have the same problem with 7 robots here 9 vision sensors.
I can’t imagine the consequences if the vision sensor doesn’t work during a race, my every step take Mobile Goal uses the vision sensor.
I can guarantee that my vision sensor is tuned very well.
I believe their issue is more with the vision sensor randomly not returning data at all, though you are definitely right that with a bit of trig you can do quite some fancy stuff with vision. Knowing the size of the goal, you could probably even calculate the distance based on the size of the readout, though that’d be a bit silly in practice probably with better solutions. Personally, my team finds using just a simple distance sensor to be best as they’re very accurate and aren’t so impacted by light changes.
@jpearman Hello.
I replaced a Brain with an older version firmware (1.0.12 and 1.0.13) and found that the problem of disconnecting the vision sensor was resolved, but the older Brain couldn’t use the GPS sensor.
So, how to use vision sensor and GPS sensor normally at the same time？
The last time we made any changes to vision sensor vexos code was in vexos 1.0.9, so I’m surprised using 1.0.13 would make any difference, however, I will setup a system next week and see if there are any issues. The gps (when not showing an image on the dashboard) should not place any higher demands on the brain than say a motor would.
I did have a look into this today and so far I’m not seeing any issues. Have a system with 1.1.0, GPS & vision sensors, been running for and hour or two with some test code just showing the gps and vision data with no disconnect issues.
So I guess I need some more details.
What other sensors do you have ? Which ports are vision & GPS in ? can you reproduce this with a simple test program ?
