It is called 'Bang Bang' from the first days of radio control in which I played a part, trying to fly aircraft. there were three positions ; neutral, or BANG full right, or BANG full left ( and they did make a bang noise sometimes, using rubber escapement motors) - no proportional in-betweens
First of all, awesome video, great explanation and demonstration of everything, i just added to my favorites! I just wanted to point out that you might've made a heavy load with that cool looking racing car model on top... and the tires/wheel you've chose does not have a good grip, its better to have something more soft to have better adhesion... another thing to consider is having bigger dc motors, those are very weak, specially with that load on top of it... please remember to take everything that ive said with a grain of salt because im not a professional at all... but i was wondering about that while i watch it run...
This is really cool. I have been researching PID for a bit now and this came up in my notifications and I'm glad it did. Great video and good distribution of humor and information. Thanks for showing the progress and the mistakes you made, mistakes are often much more valuable than the result. Thanks!
I have no idea about any of this, but I kept thinking, why not use a fast linear interpolation value to smooth out between all the hard value changes. Seems like that might act like a dampener between values and really smooth things out. Besides, I like lerp functions in all of my game dev programming. It just works great for so many situations.
@@adrianfernandes2004 Linear interpolation or LERP is taking a value and then cutting it up into a smaller part, then cutting that part into another smaller part. In programming you can do this to crate a smooth transition from one value to another that is not instantaneous. Lots of info available online, just search for "How to lerp function).
Great video, thanks for sharing it. Although I do have one question, how were you able to determine the error value accurately with only 5 IR sensors!?, I've been trying to do the same thing but couldn't see it happening with only 5 sensors. Please reply if possible
Hi im quite new to arduino but im just wondering how you got the arduino powered without plugging it in anywhere? I mean looking at 1:57 there doesn't look to be any cable from the battery to the USB or barrel plug. how is it working?
It cant take acute angles right? Im working on LFR from 2 years. Im using qtrx 16A line sensor with toshiba tb6612fng motor driver and teensy board, my lfr top speed is 6-7m/sec. Do you have any idea to crack acute angles? 5:30 your app looks similar to "Bot reboot's" pid controller app, did you made this app or using his app for pid tune?
Yes, the PID alone isnt sufficient for acute anglee. So, I coded it so that once it leaves the line it checks the previous error and turns to that side till its back on the line. This worked well on the acute angles.
it's hardly "fast", but good job on the presentation and project! (see videos from Japan's linefollowers or maze solvers, or from Europe's Robochallenge linefollowers...)
Another addition you can do to increase speed is speedup on straight sections and then slow down to normal or break on turn can significsntly boost speed
I am doing a project similar to the simple version that you show and apperantly I do everything correct but when a switch on the gears are not working. After watching your video I guess that the battery I am using is the problem. What kind of battery should I use?
At intersection point , from 3 lines which line the robot will be choose ?? How robot decided about it ?? 🤔. Is the robot will be choose a line randomly every time when it reaching at that intersection point or you programmed robot to choose a specific line from 3 lines or any other idea you taught the robot 🤔 ??
How necessary was the integral constant? Was it relatively small? Would a larger integral constant yield a better response for an error biased towards one side? Such as one wheel having (consistently) more grip than the other side?
if you use more than 2 sensors, 4, or 6 for example, with more forward view, to anticipate turns and accelerate in a straight line, is it possible? I'm a complete beginner so I'm asking this question without any knowledge...
Anticipating turns is not possible with these sensors. However, if you place the sensors in more towards the front, then it is already predicting the turns better.
Have you tried your line following robot with dead end map and does it follow the line with a mixture map(white line with black bavkground and black line with white background)?
Hello bro, first of all, let me congratulate you on the video. I'm having doubts about this line sensor. I have one just like yours, however, the sensor values always remain the same regardless of the surface it detects, whether it reflects a black or white surface, for example. I'm simply reading the sensors from the analog inputs, I'm not sure if I need to do something else?
Maybe u can integrate ann (artificial neural network) to it, so it's can learn the track, to make it some kind of intelligence enough to run even faster.
I am starting doing projects with Arduino and I am interested to start doing the simple version you show in this video. What are those components where you put the photoresistors? Sorry for asking but I am not familiar yet with electronics and Arduino. By the way really cool video dude I enjoy it!
Thank you. I have see that I will need to do it with phototransistors, are they similar to the IR detectors? Or at least if I buy this IR detectors to practice will be similar?@@aakashjana6225
How about using a QTR sensor? Me and my friends were able to make a similar fast PID based line follower, during the build we realised that the gap between the IRs was also a deal breaker (QTR being compact, solves this issue) . Also can you upgrade this to a maze solver next please, would be interesting to see that! Good video!
Thanks for the suggestion, I did consider a qtr sensor, however due to availability issues I went this one. Maze solver would be an interesting project!
I'm not sure why this was in my notifications, but this was actually made really well
It is called 'Bang Bang' from the first days of radio control in which I played a part, trying to fly aircraft. there were three positions ; neutral, or BANG full right, or BANG full left ( and they did make a bang noise sometimes, using rubber escapement motors) - no proportional in-betweens
What an evolution from the Technowright! Love your videos, I've been watching them for years!
Finally an explanation of derivative and integral error correction that makes sense.
Also, nice video overall!!!
It is unbelievable how bad teachers can explain this topic right 😂
Was thinking the same thing, if we had good and practical example while we were being taught, we would have loved it and understand so much better
Proud of you my son.... Hope to see you on another level....❤
Thanks a lot!
7:00 - At this time, you won my heart, brother! ❤
Thanks so much! It has helped a lot in my interschool project!
Its maybe the fastest line following robot I've ever seen
Great work❤
First of all, awesome video, great explanation and demonstration of everything, i just added to my favorites! I just wanted to point out that you might've made a heavy load with that cool looking racing car model on top... and the tires/wheel you've chose does not have a good grip, its better to have something more soft to have better adhesion... another thing to consider is having bigger dc motors, those are very weak, specially with that load on top of it... please remember to take everything that ive said with a grain of salt because im not a professional at all... but i was wondering about that while i watch it run...
man, i liked your video a lot, thanks for posting it. I think you could make it go super super fast
This video is incredibly good! I don't know how you still don't have atleast 100k subscribers
This is really cool. I have been researching PID for a bit now and this came up in my notifications and I'm glad it did. Great video and good distribution of humor and information. Thanks for showing the progress and the mistakes you made, mistakes are often much more valuable than the result. Thanks!
3:26 MOST HELPFUL TY CUS IT HELPED ME TO MAKE A PROJECT REPRESENTING MY SCHOOL !
Superb Shyam. Very well presented. Engaging and entertaining… All the best
Thank you!
@@shyam-ravihi cam you share how to use the code? I mean to install that somewherw? First time
Best video i ever saw in youtube.. Good explaination bro
What IR sensor are you using? please tell me specific name...
I have no idea about any of this, but I kept thinking, why not use a fast linear interpolation value to smooth out between all the hard value changes. Seems like that might act like a dampener between values and really smooth things out. Besides, I like lerp functions in all of my game dev programming. It just works great for so many situations.
And how you do that?I jus wanna learn abt it
@@adrianfernandes2004 Linear interpolation or LERP is taking a value and then cutting it up into a smaller part, then cutting that part into another smaller part. In programming you can do this to crate a smooth transition from one value to another that is not instantaneous. Lots of info available online, just search for "How to lerp function).
@@_cul8r_ dude thanks for explaining I kinda get it
Ah Yes! Quality Content❤
Nice! Great style, and very educational.
You are superhero ❤❤❤
Thanks from Egypt
Hello sir!Nice video! can you please specify what sensor you used? and does it have a digital input or analog?
Excellent video!! Hopefully we get em more often though.
واقعا عالی بود چیزی که دنبالش بودم را پیدا کردم
love from pakistan. A good quality engineering video at last 👌
Very cool , also love your apartment.
Great video, thanks for sharing it. Although I do have one question, how were you able to determine the error value accurately with only 5 IR sensors!?, I've been trying to do the same thing but couldn't see it happening with only 5 sensors.
Please reply if possible
The IR sensors give an analog value rather than a digital one. So that adds to the precision in measuring the error.
So nice innovation. Hats off🎉🤝👏
Hi im quite new to arduino but im just wondering how you got the arduino powered without plugging it in anywhere? I mean looking at 1:57 there doesn't look to be any cable from the battery to the USB or barrel plug. how is it working?
love this video , so much fun
absolutely loved this video
It cant take acute angles right? Im working on LFR from 2 years.
Im using qtrx 16A line sensor with toshiba tb6612fng motor driver and teensy board, my lfr top speed is 6-7m/sec.
Do you have any idea to crack acute angles?
5:30 your app looks similar to "Bot reboot's" pid controller app, did you made this app or using his app for pid tune?
Yes, the PID alone isnt sufficient for acute anglee. So, I coded it so that once it leaves the line it checks the previous error and turns to that side till its back on the line. This worked well on the acute angles.
@@shyam-ravi you mean, in this code it can take acute angles well?
Yes, it handles the acute angles quite well. As you can see in the video.
@@pocopoco3468
Which motors are you using
Rpm
Voltage
@@shubham9672 N20 gear motor, 1000RPM 12v
Damnn!! Amazing video Shyam, and awesome work. Loved it!!
Superb Shyam..keep rocking 👏👏👏
Very nice video shyam. Keep on rocking
Thanks!
it's hardly "fast", but good job on the presentation and project! (see videos from Japan's linefollowers or maze solvers, or from Europe's Robochallenge linefollowers...)
Fast compared to mine
Hmm, finally needed this video. Ab college aa kar batata hun :>
Finally he remembered that he also had a UA-cam channel. ❤❤
This approx. 1/10 of real super fast vids. on youtube.
really nice, what was your base speed for the motors (to be specific what was your pulse width for the PWM)
Amazing job and expalination.
Another addition you can do to increase speed is speedup on straight sections and then slow down to normal or break on turn can significsntly boost speed
The PID algorithm does that automatically
Awesome vid! ❤🔥
very nice video and presentation. You are very well articulated and a pleasure to watch.
I am doing a project similar to the simple version that you show and apperantly I do everything correct but when a switch on the gears are not working.
After watching your video I guess that the battery I am using is the problem. What kind of battery should I use?
At intersection point , from 3 lines which line the robot will be choose ?? How robot decided about it ?? 🤔. Is the robot will be choose a line randomly every time when it reaching at that intersection point or you programmed robot to choose a specific line from 3 lines or any other idea you taught the robot 🤔 ??
How necessary was the integral constant? Was it relatively small?
Would a larger integral constant yield a better response for an error biased towards one side? Such as one wheel having (consistently) more grip than the other side?
In the case of one wheel having different properties (such as better traction), you would have separate PID values for each wheel.
thanks for sharing, would you make a video and tell us the PID tuning detials?
3:00 Can I use autocad or blender instead of fusion 360?
bro's smart as hell
Could you fit a kalmar filter on the arduino to underdampen the steering?
bro what are the components that you used separately on the pcb, and which app did you use to control the pid, can you please mention.
Very nice 👍
Damn! This was awesome!
Have you tried doing it with all analog circuits, except maybe for an esp32 to let you set the tuning values remotely?
hello, can you do it with backward mode or can you guide me code with motor backward ? Thank you so much
Awesome project, and thanks for your contribution
Dang, your accent is so good
I agree
if you use more than 2 sensors, 4, or 6 for example, with more forward view, to anticipate turns and accelerate in a straight line, is it possible?
I'm a complete beginner so I'm asking this question without any knowledge...
Anticipating turns is not possible with these sensors. However, if you place the sensors in more towards the front, then it is already predicting the turns better.
Great video.
Please provide a roadmap on how we can make one of these
What to learn first?
Hi ! Do you think it is possible to implement an A* algorithm in a line following robot?
Sir which blue tooth app you are using in project for tuning of kp kd ki value.
Which charger you guys use for Charing this lipo battery? Can someone suggest any?
Have you tried your line following robot with dead end map and does it follow the line with a mixture map(white line with black bavkground and black line with white background)?
Great video!
Which app you are using for pid control
Good job bro❤❤
Bro explained PID the best way possible!!
fuck
Hello bro, first of all, let me congratulate you on the video. I'm having doubts about this line sensor. I have one just like yours, however, the sensor values always remain the same regardless of the surface it detects, whether it reflects a black or white surface, for example. I'm simply reading the sensors from the analog inputs, I'm not sure if I need to do something else?
Maybe u can integrate ann (artificial neural network) to it, so it's can learn the track, to make it some kind of intelligence enough to run even faster.
yeah we can but in some competitions this thing is not allowed
Which component you used on the pcb???
Hi ! which app did you use on the phone for controlling the PID ??
I am starting doing projects with Arduino and I am interested to start doing the simple version you show in this video. What are those components where you put the photoresistors? Sorry for asking but I am not familiar yet with electronics and Arduino. By the way really cool video dude I enjoy it!
Think he used IR detectors not photo resistors.
Thank you.
I have see that I will need to do it with phototransistors, are they similar to the IR detectors? Or at least if I buy this IR detectors to practice will be similar?@@aakashjana6225
Hi friend. Great video. What’s at J1?
bro pls tell if u dont want to use pcb and motor driver
VERY COOL, great work. But uhh, do you have the Code for the smart glasses
What is your kp and kd value that you're use?
thank you!
what is the bluetooth app u have used?
"All that glitters is not fast"
- Sun Tzu, The Art of War
i want to know your parameter Kp,Kd,Ki ,multiP,I,D is always 1 ?
How about using a QTR sensor?
Me and my friends were able to make a similar fast PID based line follower, during the build we realised that the gap between the IRs was also a deal breaker (QTR being compact, solves this issue) .
Also can you upgrade this to a maze solver next please, would be interesting to see that!
Good video!
Thanks for the suggestion, I did consider a qtr sensor, however due to availability issues I went this one. Maze solver would be an interesting project!
A very good video. Which transistor to use?
Which transistor are you referring to?
Can you tell me the list of items you used please asap
Awesome 🎉😊
Bro what voltage are your n20's
dude any details on your setup?
Bro, i thought the thumbnail was from GTA 5 Playlist race game mode. LMAO!
My brother what sensor are you using TT please tell me
what are the things on u3 c5 and c4
Totally worth it
Can you please also share the 3D design
how much rpm motor that you connected
what software is used to set the PID value?
It's been months now, still didn't got the Bluetooth control app
can u give me the pcb design u used?
it would be really helpful
Hey, you can find the link to the pcb design in my github link or instructables in description.
Is this follows LSRB algo. As well ?
how can I use qtr 8rc sensor with this code?
IS THERE A CALIBARATION AND RUN BUTTON? or just turn on power and run
First 10s calibration and run. There is no button for it
@@shyam-ravi how can i contact you bro
@@shyam-ravi bro i followed your exact tutorial, nothing happens, even bluetooth of esp32 does not show up on phone
Can I use the 5 channel TCRT5000?