Am I the only one thinking that such a system would be a hudge upgrade for diy indoor drones, without gps fixes the IMU derivates really quickly. But now it would need to account for varying altitude, pitch and yaw. This system is already implemented into DJI's and parot's drones with bown facing cameras, and having an accessible platform would be really appreciated
Can the optical sensor estimate the position/displacement for a flying drone? For robots that are on the ground, the optical sensor is close to ground so the optical sensor should work. Think about the case of an optical mouse not working when you raise it.
@@TahaKoroglu I recognise that it would become way more complex for a flying platform, but with some lenses you could adapt the sensor to a range of heights, and with the imu inside, you could account for the pitch and yaw of the sensor. It wouldn't be easy, but as such there is a market for a working sensor
Many drones already use so called optical flow sensors for precision tracking. All current DJI drones use them to hold their position more accurate when close to the ground. This is how they can hover very accurate close to the ground. This is also how they hold their position indoors, where gnss is unreliable. Additional TOF sensors and Barometers help with holding altitude. Optical Flow sensors are limited in range though. The ones available for DIY use state that the work within 8m off the ground.
Yeah they make a ton of these. For example the matek 3901 optical flow sensor is essentially the same thing made specifically for drone software like betaflight, inav, and ardupilot. The one thing you'll find different from this though is that all the drone optical flow sensors must have some kind range finder to make the calculations work. Barometer isn't really good enough in this case.
Oh my god this is awesome! I've been looking for an optical flow sensor from a mouse for exactly this purpose! You can't buy them though. I was never able to find the actual sensor for sale, it seems like they all go directly to mouse manufacturers. This is so good!
Hi! As a robot builder for the French Robotic Cup, and since we build holonomic robots, I ordered this sensor on day one I just received it and I'm experiencing some issues with the IMU. It constantly drifts, even after calibration. That’s sad because all the other features work like a charm. But this drifting on the angular position destroys the measurement accuracy. It's a shame because otherwise, it’s a great product. Perhaps a different calibration method or software corrections are needed?
at 8:02 he covers how to tune to minimize drift on the angular measurement if u did that and it still messes up and u think its a device issue (which it might be given you bought it very early), it might be worth looking for a replacement/refund on the sparkfun site
Hi Sparkfun, I'm getting in trouble trying to find how to wake up CO2 sensor STC31 when using your library " SparkFun_STC3x_Arduino_Library", can you tell me which function to use to wake it up after enter sleep mode
looks like you could do a 'next-gen' version of this with STM's intelligent IMU, that can perform a LOT of navigation tasks on-board....like ISM330DHCX.
Yes, of course, I'm also a TH-camr, but nothing says that you fixed the subtitles. And speaking slowly and clearly is an excellent exercise when addressing the whole world.
@@enginerdyif I want text, I'll buy a book. Not only does he speak too fast, he swallows common specialised words. I'll get to a, specific example in this video, but how many times, do we hear socially involved people saying comuniddy, or worse comune-ee? I this video I2C (I think) was virtually unintelligible.
Woot! Go FIRST Robotics! FRC Alumni and Mentor here! This looks to be a fantastic solution for many FIRST teams in all the different competition
This is great; I really appreciate having an/the engineer speaking to the product and their experience and insights
Great guy, great voice, great pacing. Cool practical example! Great to see the SWD pins broken out and Firmware available!
too fast...
Awesome! I also had the idea of using a mouse years ago (I'm an FTC mentor), so glad to see this become a reality.
Am I the only one thinking that such a system would be a hudge upgrade for diy indoor drones, without gps fixes the IMU derivates really quickly. But now it would need to account for varying altitude, pitch and yaw. This system is already implemented into DJI's and parot's drones with bown facing cameras, and having an accessible platform would be really appreciated
Can the optical sensor estimate the position/displacement for a flying drone? For robots that are on the ground, the optical sensor is close to ground so the optical sensor should work. Think about the case of an optical mouse not working when you raise it.
@@TahaKoroglu I recognise that it would become way more complex for a flying platform, but with some lenses you could adapt the sensor to a range of heights, and with the imu inside, you could account for the pitch and yaw of the sensor. It wouldn't be easy, but as such there is a market for a working sensor
Many drones already use so called optical flow sensors for precision tracking. All current DJI drones use them to hold their position more accurate when close to the ground. This is how they can hover very accurate close to the ground. This is also how they hold their position indoors, where gnss is unreliable. Additional TOF sensors and Barometers help with holding altitude.
Optical Flow sensors are limited in range though. The ones available for DIY use state that the work within 8m off the ground.
Yeah they make a ton of these. For example the matek 3901 optical flow sensor is essentially the same thing made specifically for drone software like betaflight, inav, and ardupilot. The one thing you'll find different from this though is that all the drone optical flow sensors must have some kind range finder to make the calculations work. Barometer isn't really good enough in this case.
I have a couple of optical flow sensors and an IMU but packaging this on one board **and** doing the sensor fusion for me is well worth the price!
WELL BACK IN MY DAY...
crazy stuff. im hoping this becomes more well known and turns into a major contender for the best odometry
FTC Rocks! (Former FTC judge and team 5339 mentor)
Oh my god this is awesome! I've been looking for an optical flow sensor from a mouse for exactly this purpose! You can't buy them though. I was never able to find the actual sensor for sale, it seems like they all go directly to mouse manufacturers. This is so good!
Hi! As a robot builder for the French Robotic Cup, and since we build holonomic robots, I ordered this sensor on day one
I just received it and I'm experiencing some issues with the IMU. It constantly drifts, even after calibration. That’s sad because all the other features work like a charm. But this drifting on the angular position destroys the measurement accuracy. It's a shame because otherwise, it’s a great product. Perhaps a different calibration method or software corrections are needed?
at 8:02 he covers how to tune to minimize drift on the angular measurement
if u did that and it still messes up and u think its a device issue (which it might be given you bought it very early), it might be worth looking for a replacement/refund on the sparkfun site
Hi Sparkfun, I'm getting in trouble trying to find how to wake up CO2 sensor STC31 when using your library "
SparkFun_STC3x_Arduino_Library", can you tell me which function to use to wake it up after enter sleep mode
👋 Hi friends!
Great video. I think it would be better to slow down your speech a tad bit, put otherwise, great job! SparkFun rocks!
x0.75 playback speed
looks like you could do a 'next-gen' version of this with STM's intelligent IMU, that can perform a LOT of navigation tasks on-board....like ISM330DHCX.
Has any FTC team tried this out yet?
Good stuff but the price needs to come down.
Compared to 3 odometry pods, this is a great price.
Neat!
Cumulative error 🤯
Excellent product. Please speak more slowly for those whose native English is not their first language.
Closed captions appear to be quite accurate
Yes, of course, I'm also a TH-camr, but nothing says that you fixed the subtitles.
And speaking slowly and clearly is an excellent exercise when addressing the whole world.
@@oracidI play all videos at x2 speed, you can reduce the speed if you prefer
Thanks a lot for this tip. It works very well. I reduce at 0.75 and it is very good. He should speak that speed.
@@enginerdyif I want text, I'll buy a book.
Not only does he speak too fast, he swallows common specialised words.
I'll get to a, specific example in this video, but how many times, do we hear socially involved people saying comuniddy, or worse comune-ee?
I this video I2C (I think) was virtually unintelligible.
How well would this work outside on concrete in the dark?