Accelerometers and Gyroscopes - Sensor Fusion #1 - Phil's Lab #33

[@YoursTruelyMe2]

Man, this corner of KZbin right along with Ben Eater and 3brown1blue channels are among the best

[@ianmosquera3741]

I agree, these three channel is the holy grail.

[@ruben.w]

Love the extra effort in addressing the drifting problem not only in a theoretical way, but showing this in an (un)practical scenario.

[@PhilsLab]

Thanks, Ruben - very glad to hear that!

[@helgeb5403]

Your channel is a pure bliss. Profound and still condensed knowledge. Are you planning on doing another video on FIR and IIR filtering including the FMAC peripheral of the stm32 MCs?

[@PhilsLab]

Thank you very much, Helge! I haven't planned any videos on the STM32's FMAC yet I'm afraid..

[@Ngadakwiyokwiyo]

You're life changer Engineer Phil, I appreciate your wonderful contents,

[@MathewPanicker1010]

What a happy coincidence! I had just started to look for educational content on sensor fusion this week.

[@PhilsLab]

Awesome! Thanks for watching :)

[@lmmlStudios]

mate this channel was an instant sub in the first couple minutes

[@wizardOfRobots]

The demos were great. Usually only the theory is explained. Thanks. Looking forward to the next videos in the series.

[@PhilsLab]

Thank you very much!

[@brus54per]

Really nice presentation, thank you! Sensor fusion is a wonderful rabbit hole where one can spend all time until retirement if need be ;) It would be very nice if you would include quaternion-based solutions as well. With modern processors, that is a viable route that offers some very interesting possibilities. Good luck and happy fusioning!

[@Andres-is8zz]

+1 to this comment

Love the video and looking forward to the next part with sensor fusion, I have not yet managed to wrap my brain around Kalman filters.

[@PhilsLab]

Thank you! Hopefully the Kalman filter video can clear that up a bit :)

[@davidegandolfi25]

I can't wait for the Kalman filter video, because I'm aware of it, and of sensor fusion in general, by quite a long time now.. however this may be the first time that I really understand it, because your explanations are really clear and to the essentials!! Thank you so much for your work!

[@vintyprod]

Thank you for these. The quality of information is incredible.

[@denysvisser]

As someone that does not work in but adjacent to this field these videos are amazing at building knowledge to better communicate and understand this stuff. It really is a gem! Thank you!

[@user-qf6yt3id3w]

I like the way you keep the math as simple as possible but no simpler.

[@PhilsLab]

Glad to hear that, thank you!

[@iamnarval]

This is a super welcome video, thanks for the effort! How about a 4th part too with quaternions? :)

[@PhilsLab]

Thank you, Peet! Good idea, I may add a bonus quaternion-based EKF as a last video :)

[@DrGreenGiant]

Came here to say the same thing about quaternions. This is what we did for an aerobatic UAV to get around the Euler issues at 90 degrees

[@sydmustbehave]

Thank you Phil for this informative content. Excited for Part 2!

[@abhishekreddy2425]

THIS IS AMAZING!!! Looking forward for the next video and super excited!!!

[@danimal_1814]

Excellent video! Thanks to You and your sponsor.

[@jeffcarter4500]

So happy I found your channel! Thanks so much, keep up the great work!

[@iwbnwif]

Really great series, this is such a useful resource for some IMU experiments I am planning!

[@MrZomhad]

Exciting content as always! Looking forward to the next videos of the series :) Also really enjoying the slides!

[@PhilsLab]

Awesome, thank you! :)

[@bhu1334]

Thank you for this video, im greatly looking forward to this series. Have you looked into the Madjwick filter, it seems like it's less computationally expensive than the EKF Also do you plan on making a video on integrating the attitude estimates with GPS?

[@PhilsLab]

Thank you! I've played around with the Madgwick filter but wasn't happy with the performance and expandability in comparison with an EKF, although it is less computationally expensive. I'll probably add in IMU-based GPS-smoothing in a future video (not this series however, as this'll just cover the basics).

[@microcolonel]

What I'd like to see more of is fusion with a structural model of the vehicle and MEMS combos at multiple points on the structure.

[@abhijitkamble4519]

Awesome content... really appreciate your efforts. Thank you

[@danielye9544]

At 10:22 I think Phil meant to say true angular *velocity* (not angular acceleration).

[@MikeNugget]

Awesome video! Can't wait for the next.

[@PhilsLab]

Thank you, Mike!

[@mikegofton1]

Great content Phil - looking forward to the filters video. It would be useful if you could include commentary on additional sensors (e.g. magnetometer for yaw , thermal horizon sensing for pitch and roll).

[@PhilsLab]

Thank you very much, Mike! Yes, I'll touch on using a magnetometer for heading estimation when we come to the EKF!

[@sandman94]

thank you for the detailed explanation 👍

[@ariswidiyawan3323]

cant wait for part 2

[@sarbog1]

This is very cool!!!

[@PhilsLab]

Thank you, Stephen :)

[@sarbog1]

@@PhilsLab Please feel free to go into more of the mathematics... Love the combination of Physics and Electrical Engineering.

[@chinoramas1]

A very informational video, Thanks!

[@PhilsLab]

Thanks for watching!

[@thiagovs.s]

Amazing channel!

[@PhilsLab]

Thank you!

[@abder5453]

good job man .. keep it up

[@PhilsLab]

Thank you very much!

[@sudharsan3835]

Thanks for the video. This is gold.

[@PhilsLab]

Thank you for watching!

[@EmbeddedEnigma]

bless you for this channel and in this video. Keep em vids coming learn a lot from them

[@PhilsLab]

Thank you very much, Haseeb!

[@myetis1990]

It is quite nice to see somebody hearing me ! (I've cited this topic in the previous posts ) great content and very appreciated! Btw if you enlighten the gimbal lock issue after this fusion topic, it will be very appreciated. thanks in advance.

[@PhilsLab]

Thank you, Mustafa! :) Yes, exactly - we'll look at the gimbal lock issue in the next two videos.

[@mostafakh5075]

that's what I'm working on it these days, great 👌. I'm implementing imu to achieve yaw. i set it on my desired point and set it to zero, then when i rotate that it gives me a good yaw at the first rotation but after that it start the random walk and drifting the system, i don't know how to solve it. in this case i dont use magnet

[@Phil659]

Awesome content, thanks phil

[@PhilsLab]

Thanks for watching!

[@MsKengster]

Great video! I am really interested in learning all about IMU and all the implementation methods. I would like to really understand how everything works but even If I am an engineer I feel that I need to re-study everything again. Could you please recommend me some good technical books for learning about this? Thank you!

[@dansam1395]

Great theoretical and practical explanation :) ! do you practical advantages in using quaternions for attitude estimation?

[@MrRonychakraborty]

Yes waiting for ext kalman filter :)

[@PhilsLab]

Thanks, Rony! :)

[@alexlo7708]

Time varying bias term -> drift.

[@Retinatronics]

Interesting topic!

[@PhilsLab]

Thanks!

[@DonQuichotteLiberia]

Excellent, thank you! Are you planning to touch upon positions and velocities (e.g. from GNSS) too?

[@TrungNguyen-wq5kw]

The formula you used at 10:44 was seen in many articles, all of which uses “plus” quadcopter setup to mathematically model. But I always wonder if plus setup and X setup would be the same, or if I could use their result for an X quadcopter. If not, why use plus, while X is more practical.

[@hristiantodorov3923]

Great video, Sir, thank you! Why did you have to inegrate the Euler rates, you already had the phidot, thetadot ?

[@eledikohabib3369]

Please consider making this on the RP2040

[@ruffy4004]

Great content!

[@PhilsLab]

Thank you!

[@mcgdb3606]

awesome lecture... what's that serial oscilloscope you're using?

[@PhilsLab]

Thank you! It's this one here: www.x-io.co.uk/downloads/Serial-Oscilloscope-v1.5.zip

[@gasqui]

I'm getting a NaN when acc_x is greater than 9.81, because sin^-1 of something >1 is a Mathematical error, so I was wondering if it is due to my accel sensor reading or maybe the pass low filter or I have to declare a constraints? Btw, terrific tutorial, thank you a lot.

[@lalinlalote]

Do you recommend a book with all these topic in this amazing platical way?

[@mounirdahlal5350]

Do you have books in this field?

[@soulrobotics]

...you just put my signals and systems professor to shame in 14 minutes...

[@rick_er2481]

Awesome!

[@girayyillikci3188]

thanks mate

[@sudayshankar9036]

Could you make a video on pcb design of nb iot modules

[@mohammadaghazahiri2456]

auto transcript is set in German language! Can you please fix it?

[@navyblu5064]

can i have the references for the IMU's model?

[@gillsejusbates6938]

Hello, your handwriting of Theta is a crime against Greeks!

[@konturgestalter]

loooove it

[@PhilsLab]

Thank you very much! :)

[@lhxperimental]

Watched the entire video for sensor fusion only to find at the end that sensor fusion is in the upcoming video. 😭

[@legendarycraft5499]

+1 sub :)

[@deviljelly3]

Well that was easy..... 8-/