Controlling Self Driving Cars

[IEEE CSS Video Clip Contest 2015 Submission]
This is a video introduction to controlling self-driving cars, specifically using PID-control. Special emphasis is placed on how the proportional, integral, and derivative gains affect the performance of the vehicle.
Music:
www.youtube.co... - Solar Flares
Video References:
DARPA Urban Challenge MIT footage
Google Earth
Dog Driving Car • Meet Monty. The World'...
Car in Parking Lot • Strong Wind Blows Over...
Truck • Video
Credits:
Script & Narration: Luke Johnson
Hardware & Control: Brett Lopez
Visualization: Shih-Yuan Liu
Editing & Overlays: Justin Miller
Lab footage courtesy of:
Aerospace Controls Laboratory @MIT
Jonathan P. How, Director

[@illusivec]

As someone with masters on control engineering, I can safely say that this is the easiest to follow introduction to PID controllers I've seen. Wish my professor had told it like this in the first class instead of throwing 4 whiteboards full of Laplace transforms.. .

[@mj1234321]

Haha, couldn't agree more!

[@Dra60oN]

same here

[@jonasjo93]

So true

[@valdisgerasymiak1403]

+! dude)

[@kabascoolr]

Same here. My mouth was agape half-way through.

[@kamashi85]

PID beautifully explained.

[@znx418]

Couldn't agree more!

[@jobsmine]

By far the best PID control tutorial i've seen. Far better than that Aero engineering class I took in last fall term.

[@freelectron2029]

i disagree. kzbin.info/www/bejne/p3zcamt5l8ifqtk

[@Sleeperknot]

This is probably the only video that gives a proper intuition about PID control. I wanted to develop that intuition to tune the roll, pitch, and yaw controls of my drone. None of the drone experts know how to properly explain the impact of changing P, I, and D.

[@Nick-kb2jc]

Currently working on implementing self-driving vehicle behavior for my video game, and this is exactly what I needed! Thank you so much!

[@mainsequence1055]

One of the best explanations and associated animation of PID control I've ever seen. Well done.

[@freelectron2029]

i disagree. kzbin.info/www/bejne/p3zcamt5l8ifqtk

[@thunderloopthriller]

Thank you for this very clear explanation! I have been trying to get my head around this concept late at night, and your explanation helped me (and also allowed me to relate it back to other learning about damped oscillators). I'm working on a simulation with numerous control systems, with only proportional gain, and it's super finicky to tune. Thanks to your video I now have a better idea about why this is so, and also know what I can do about it.

[@hadijaffri9856]

the best PID algorithm tutorial I have seen. The visual aids makes it so much easier to understand and give an intuition

[@GranVlog]

And our brain does this perfectly once we know how to drive. now that's cool.

[@ajelvani]

Our Brain works more like an MPC (model predictive control) I think

[@gasun1274]

our brain does not use the PID model, although for some tasks we have been observed doing this

[@4ae109]

our brain mimics this for controlling our limbs.

[@hooblymcdingelbary1311]

Found this while trying to understand PID in 3D printer temperature control. Amazed at how almost exactly the same these issues are

[@NoajmIsMyName]

I took this class and I just now know what does PID mean. Well done my Prof "You know what I'm saying" I heard that sentence more than the times I heard my name

[@chimingito]

god damn I finally found a video that properly explains every part of a PID controller.

[@abdullatesttest6472]

PID BOOOM! As I watch more and more videos and read more on PID control, it is getting clearer to me! fascinated by this beauty!

[@FPVSpace]

As an FPV quadcopter flyer I found this very interesting. We use PID values to tune our quadcopters.

[@GranVlog]

+FPVSpace right on

[@frodo_banggin]

+FPVSpace agree

[@caturlifelive]

Agreed, up

[@freelectron2029]

i disagree. kzbin.info/www/bejne/p3zcamt5l8ifqtk

[@SurvivalSquirrel]

BEST video ever for PID! Even under 5 minutes. Good job!

[@freelectron2029]

i disagree. kzbin.info/www/bejne/p3zcamt5l8ifqtk

[@KhoPhi]

I came for the example and explanation of the PID, but oh dear, this caught me off guard. What a recovery that was! A human PID tuned to perfection! 3:02

[@marcelosantospires4064]

Im a brazilian in high school and im preparing to a robotics competicion, thank you so much, this explanation was perfect to me.

[@sethadkins546]

FRC?

[@devangjoshi1194]

Very good explanation on PID controllers

[@TheEarlyAstrotype1024]

This is currently the best example of PID control I have ever seen.

[@Lergurker]

The best PID explanation I have sen so far! Good Job!

[@aaqilkhan]

Very good introduction to closed loop PID controllers. Video and narration was very well done.

[@lawofrobotics]

This is the best control theory video I’ve ever seen.

[@maxim25o2]

The best explanation and visualisation ever. No body else can do it better.

[@MrsBangladesh1971]

the best tutorial I have found so far on PID control. Thank you! it helped a lot.

[@freelectron2029]

i disagree. kzbin.info/www/bejne/p3zcamt5l8ifqtk

[@hfe1833]

Of all all youtube video i saw regarding car autonomy this one is the most detailed and easy to understand

[@philnguyen834]

I can say this is the best explanation of PID control on Utube

[@aktunav]

One of the best PID explanations that i have ever seen...

[@EvillNooB]

very well explained, i love when the complex topic is explained via breaking in down into simpler pieces

[@Tavnott]

That made way more sense than other videos! Thank you so much for your clear explanation!

[@Extner4]

Props for them to use an actual rc car, trackers and a projector instead of just a simulation

[@frillneckedlizard8529]

this is exactly how I created my ksp throttle hover control script, And I didn't even think about it that much

[@boltonky]

I feel like i am going to forget this cause its late, but one of the best and simplest explanations i have seen....when you learn things the smart way :) sucks we all learn different at times

[@Tunquenino]

This video with the formulas are the perfect example of the new way to learn. Nice videos guy!

[@TIHutch1]

That Save tho! a@ 3:03

[@rdsully1221]

this is exactly what i was looking for. Im in design school prototyping a autonomous beer buggy

[@BoomZaTvZ]

The best PID explanation

[@bpinaya]

This video is beautifully well made. Awesome job guys!

[@ksthar5631]

As expected from MIT :) so beautifully explained about the PID ^.^

[@muhammadhabibalkhairi7946]

Really good explanation about PID without mention PID first.

[@tigerexo6577]

It’s always a amazing how our brains can function so well & handle so many things.

[@ethanoch]

Fantastic video! A great introduction to PID control.

[@AlqGo]

Your video explanation is intuitive. Please make a video on Model Predictive Controller, Linear Quadratic Controller, H-Infinity Controller and Observer, Sliding Mode Control, Controllability, Observability, etc. :D

[@akhil6095]

One of the best PID controller explanation!!!

[@Bloomio95]

How to click the thumbs up multiple times?

[@JBSessionss]

Wish I've seen this like 4 hours ago, before my exam on sensors.

[@ocieward]

Very well stated and simple explanation of PID control. Kudos.

[@fuckchannels5609]

Pro truck-driver!

[@andreagiudici926]

one of the best explanation of PID control!. thanks!

[@RJ_RJ_RJ]

A great explanation, bravo

[@jakubtomasek]

Very nice visualization!

[@FayyazeEbrahim]

That's the most beautiful lecture, I have ever heard!!

[@jaswanthayrus]

Sorry, but I couldn't get what's going on at 3:19 . I mean, even though the car gets off the track due to some external factors, knowing that it is off the path, the already existing P and D are there to steer it in the right direction right?

[@REL1C]

Great video and explanation on PID control. Thank you.

[@alhdlakhfdqw]

incredible! thank you so much for your wonderful video! pls do more! subbed

[@Must23]

The best explanation ever

[@BadPractices]

What a phenomenal video.

[@MePeterNicholls]

I’d intuitively reasoned out these three ideas from analysing how my car drives and holds its lane centre, but only in a basic form and had no real grasp of what was at play. This video was brilliant, explained it really well while taking us through to an understanding of the higher concepts and mathematics!

[@thinkbig899]

Awesome nice excellent

[@navaneethmnambiar2577]

Great video and explanation of PID

[@sauloroncon]

good explanation on PID. Congrats guys

[@phillipnewman9174]

Brilliant! An excellent presentation! Easy to understand.

[@dineshlamarumba4557]

Link to code?

[@ithinkyouforever]

great job for controll PIDs thank you

[@ChillGuy511]

Brilliant video!

[@bunyaminesiyok8519]

Nice

[@justingt3rs]

# passionate about your thesis

[@kaden56]

It’s to bad most professors are too prideful to show something like this day one of classical control theory. If you start with understanding it from this simplistic yet very accurate view point then at least all the heavier math is pointed somewhere.

[@abdotsouri9432]

how i can find the best value of the constante

[@LukTzeChing]

nice video, sadly that my professor doesnt show this kind of visualizing video in his lectures, otherwise I wont get a B grade and skip almost all of the class

[@santhoshmamidisetti]

Can you please name those advanced controllers u mentioned at 4:10 please

[@shaflyhamzah3848]

Nice explantion!

[@الحمدلله-ظ8ب6ظ]

I want the code that do this please

[@hnl5100]

for those who want to implement this in python a code here you have the simplest way to do this: class PIDController: Kp Ki Kd error_sum = 0.0 error_rate = 0.0 _prev_error = 0.0 def __init__(self, arg_p, arg_i, arg_d): self.Kp = arg_p self.Ki = arg_i self.Kd = arg_d def update(self,target,current): error = target - current self.error_sum += error self.error_rate = self._prev_error - error self._prev_error = error return self.Kp*(error) + self.Ki*(self.error_sum) - self.Kd*(self.error_rate)

[@ufuf888yg6]

Pretty nice explanation!

[@Excalibur32]

Excellent video, thank you for sharing!

[@abbymerchant2069]

huh thought it was gonna be a video on MPC but ok i guess its a quick PID vid.

[@jannuellaurodizon3016]

This is so good, damn

[@GeezerRay]

Pretty standard control-system theory.

[@John_Ridley]

+Ray Perkins Yes, and there are a lot of videos explaining it, but fewer more clear than this. IMO you can't have too many explanations of such basic concepts, because you want as many ways to explain them as possible to maximize the number of people who really "get it."

[@cariyaputta]

Title nothing mention PID, but PID beautiful explained.

[@arialuo3259]

Omg I finally understand this! This helped a lot thanks!!

[@AdityaSingh-pj8ie]

This did a great job explaining the concept, but I'm still confused with the Integral term. If the car's spent time off the track then the integral term increases and drives the car back to the line. But once it's on the line, if P and D are zero, aren't you still left with the I term pulling the car away? At 3:43, the car moves along the line perfectly horizontal, but both the I and the total vectors point down. Is the I term supposed to be reset to 0 when it hits the line?

[@shravanrajagopalan4387]

Aditya Singh that's a really good observation. It's precisely for the same reason that the integral gains are not set high as it could cause the car to overshoot. However if the I gain is tuned low, then the derivative gains will resist the changes caused by the integral gains.

[@jsquaredm]

I have never had a class on PID controllers, but I am trying to use them in my holonomic robot. It is really difficult trying to tune a PID controller, while dealing with the motor configuration of a holonomic robot!

[@stevenchou5905]

I love this video, and I love you

[@flavianonzambi6878]

amazing vid

[@overPowerPenguin]

Amazing explanation.

Great content

[@macmaniac77]

Love this!

[@33athen]

could i get more details on the implementation of PID

[@tecsupcesarmeza5741]

Amazing Video!!!!!

[@RishuKumar016]

Can you share the code and design requirement? It will be of great help

[@SezerYalcin]

You should use car body approach angle in proportional gain method and add steering wheel angle to control that parameter. Your approach to relate steering wheel angle directly is wrong.

[@bsingh9653]

you are great

[@manavshah8239]

Great explaination for intermediates! Keep it up!

[@okaythen2295]

Thanks, great explanation

[@kuisongzheng4655]

what is the cross track error to a curve? will the car follow a curve path?

[@faizakram3686]

Which microcontroller is best for this type of control??

[@angcongbui8508]

Thanks for this great video. I just have a question, how can we derive the relationship between the cross tracking and the steering angle? Are there any documents written about this? Thank you in advance.

[@الحمدلله-ظ8ب6ظ]

By code use open cv paython

[@rgloria40]

So what is the back up system.....Embedded road sensors?

[@eufenico]

Outstanding!

[@azizjamal-de8ti]

This is the best