2014 APEC(28th) Micromouse Contest Final, USA

[@douro20]

I think the most technically impressive one was the one with the built-in camera which used images taken from several angles to find the most efficient route to center.

[@sanjaybarai8555]

douro20 ßùmjķďgnh utre

[@jacobnielsen5319]

I would like to see a micromouse spilt into 2 or more sub mouse to explore the maze faster by communication their discoveries with each other, and then work together to get all mouse to the finish goal

[@Greenye]

Jacob Nielsen u think it will against the cureent rules by leaving stuff behind. Lol.

[@pawelec555]

just make them connect back up :d

[@SetTheCurve]

@@pawelec555 you could have a spool of. Wire that connects the heavy batteries with a forward section that completes the maze. Still attached :)

[@marvinochieng6295]

that is a brilliant idea

[@nikolaswithak]

The Japan one was amazing.

[@anakinskywalkerrr]

just because its different, which is good

[@mister_nxt5442]

Yeah it was cool. Also it was clever that used a picture instead of just exploring about the Maze. Also for convinemce heres the timestamp. 4:58

[@proke7855]

@@mister_nxt5442 love you mate

[@B34RD]

But can they vacuum floors?

[@1dan1609]

I watched this video 3 years ago, and searched the whole KZbin just for that camera one.

[@King_Kristof]

link or it didn't happen

[@Scwirul]

I wondered why it seemed the 'fast version' of each run was counted as points... not necessarily (or entirely) the first slower run... then I realised that the mouse was learning the maze, and as it made it's initial way back from the center, it was trying to find an alternative faster route! So the initial slower run wasn't necessarily the same route as the second faster one... That was pretty interesting to me!

[@Greenye]

because the best score = current run time + (past maze time /30) + possible penalty. Therefore the best score is likely to be the first speed run without touch penalty.

[@repely]

Oooohhhh I get this now

[@noriorck4017]

Even though the camera bot lasted a long time procesing the path he used only one run to learn the maze and run it, wasnt he suppousted to win?

[@benjimoomoo]

the way these comps work i think, is that the first run is used to 'learn' the maze, then the second run is all about speed. the problem with the camera bot, is that is as fast as it goes. all the others went much faster on their second run.

[@7xr1e20ln8]

@@benjimoomoo since he never used 2nd turn, doesn't that mean 0?

[@vincentarisa9513]

Exactly what i thought that was out of the box thinking

[@NicD]

Love the Mario victory music by the 4th place! 14:57

[@DF-wo5lb]

for me, he of the camera is the best :)

[@Fighterbharat]

David Figueroa the problem with camera bot is in actual caves or holes you cant cheat by taking a helicoper shot of the fuckin maze.

[@SilverforceX]

@@Fighterbharat Yes you can, give it a sonar or ultrasound camera to map out the maze through rock.

[@НикитаБарабас]

Японцы. Как всегда на высоте. Все гениально просто !

[@BayMacDre415]

I wonder why some robots would pass up the first right turn @13:11 but then successfully hit the second right @13:19, instead of going to the end of the hall and going around the corner?

[@lastchance1036]

Some sort of optimal algorithm I assume. If you pay attention it also has some unexpected turns throughout the runs

[@dreamEnd]

so what determines the ranking placement? is it time?

[@joaoalvesdesouza2812]

Vocês são feras ..parabéns.🇧🇷

[@joshuagallup3005]

I was expecting actual mice...

[@memesmfiaz1607]

Hehe I was watching mice in narrow maze then saw world competition..I thought wtf there's actually a sport in this....then I see robots...hahaha

[@EastAtLeast]

THE FUTURE IS NOW, OLD MAN!

[@jun-ob6sk]

East no, now is the present...

[@matador6604]

imthiyaaz fiaz was there a shrew involved 👀

[@SpaceCircIes]

the robot knows where it is at all times, it knows this because it knows where it isn't

[@CCheukKa]

How are the mice scored? The rankings make little sense to me

[@Luca-pd6bq]

As far as I know fastest run from start to finish. They have a set amount of time for exploring and trying to finish as fast as possible. Or they have two runs were the first is just for exploring and the second is for taking the time.

[@CCheukKa]

@@Luca-pd6bq I feel like some lower-ranked mice have faster completion time than some higher-ranked ones?

[@AntonioVivaldi1678]

Man, just gotta love Fort Worth Taxes. 0:00

[@strutter789]

Can anyone explain the reason for the lint roller around 8:30

[@Greenye]

clean the tire with better traction for speed run.

[@davideophile]

01:10 it usually.... What?! tell me! WHAT?!

[@fsofarda7871]

that camerarobot is my winner

[@bummers]

Just before this video, I watched one from Japan. It finished in 7+secs. You barely see the micromouse ... it is like 1/3 the size of the "micromouses" in this video.

[@Hellefleur]

I wanted to hear the announcer talk about history and trivia. I'll have to see if there's a full version anywhere.

[@jacobruiter592]

I liked the super efficient and messy one that kept beeping while running into walls. I was still programmed better in the sense that it knew the end point. It was also SUPER FAST. In my opinion the ones that do not know the end point need work to be perfect.

[@Joshuaxiong2]

8:16 Why no one clapped?

[@ramsie50mp]

Review rules of entry: Camera telescope breaks the plane of maze. the optics are supposed to be from top wall to floor not above.

[@dl950]

Why can some know the path and others not? Are they allowed to reprogram after seeing the maze?

[@Greenye]

they are not allowed to reprogram the mouse. no one knows the path before search. The camera one took a picture of the maze and did some image processing.

[@kaeyaalberich9594]

2:11 is the best the robot said "AIGHT IMMA HEAD OUT"

[@nikolaswithak]

Does at the beginning it mean "Fort Worth, TEXAS" not TAXES

[@Greenye]

You are right, but too late to change :-(

[@Tedisdeaad]

You can still fix it in the description though

[@Greenye]

too lazy XD

[@treyguffey9560]

Green ye saaaaame

[@landenbrown3232]

How do you misspell Texas?

[@Greenye]

yeah, shame.

[@kjb21161]

How is this scored? cause the 2nd place was way slower than 3rd???

[@OttoByOgraffey]

The injectifier in the trombolinier with subfractors interdispersed throughout the varience of the quadmidifier for the 4th mouse (mus) , should invariably, by a common denominator, besmurchified the overkinetic significance between the trachiadacotimer upon the 4th power without significance, unless, of course, it is without the partisimonly. Almost every time. I was wrong once, but then I was mistaken.

[@Litepaw]

Otto By Ograffey 'scuse me?

[@masonhmusic]

Otto By Ograffey huh?

[@OttoByOgraffey]

Mason Hunn, congratulations! We think the same. Very good.

[@darioildiario3001]

You forgot the redundance of the post gradient analytic synopsis: the polarization of the over-clip has more significance, if you consider the partial predominance in the structural framing... geez

[@OttoByOgraffey]

dario il diario , no, no, no. You can't have it both ways. This has been proven time and time again. The under/over psumpler, the p is silent, is in contrast to the polarization magnums by 30. It's been proven. SMH. I thought everybody knew that.

[@Keyakina]

What kind of sensors do they use? I mean with ultrasound or something they could map way more instead of driving against a direct wall?

[@Greenye]

Sjard Willems v infared sensor. Encoder, gyro

[@asherael]

how's the mitee13 know the way, was it after a training run?

[@Greenye]

it performed a search run first. It was included in this video.

[@asherael]

@@Greenye i must have missed it. Thank you

[@RareBrink]

fascinating competition, I loved seeing the bots that were just complete speed demons. The camera bot is cool but all that extra weight really lowers his overall agility.

[@rembrandtes]

thank you green ye .i like your videos, they are great as usually

I wonder who would have fastest time without second run. A first run for mapping the maze is nice idea, but kinda beats the purpose.

[@Itachi-wk2ju]

Can I now what are the programme used In that robotic car...

[@プロテイン岡田-h4i]

15:42 Japanese one

[@hodor6554]

Duuude It took me 3 days trying to assemble my shoes cabinet and i ended up calling the company to do it for me

[@MassDynamic]

2:13 so close, yet so far.. the japanese guy was pretty clever to use a motorized camera

[@coolio227722]

Okay why did the camera robot not win?

[@stevesoto462]

How did they grade them ?

[@planktonfun1]

why are only old people participating?

[@John-mj2ct]

Paulo Castro because they're smart

[@planktonfun1]

and have jobs to make the bots, jobs for younger people should be legal

[@pupzakuro3300]

Paulo Castro do u read before you post?

[@planktonfun1]

PSC Zemark556 I guess you're thinking slavery. Common people would assume that, but for others like smart kids to unlock their potential they would usually need money that their parents cant provide

[@wtfftw24]

Wasn't the second faster than the first one?

[@xdee4502]

22:33 awkward handshake

[@dannyhercus6379]

Is the announcer AVE?

[@ExopMan]

So the competition results are only based on the second run? Once the mouse has scouted the maze?

[@21yashthakur]

22:33 handshake failed😂😂😂

[@mikk5802]

2:08 That's me when i forgot to buy milk.

[@DanialADHOfficial]

what course in university to study this thing?

[@jasonhurtadoo]

ING electrónic

[@jakea7065]

Are those bots made on custom PCBs?

[@Greenye]

Jacob Allison yes

[@skytaniguchi1198]

how do they judge these because some of them should have gotten better ranks then others

[@Luca-pd6bq]

The fastest run from start to finish in a set amount of time. As far as I know.

[@ramones53rd3rd]

The one that cheated and took the picture first was dope lol

[@tsuyoshimorino2602]

Maze which can be solved by sticking on either hand, yes but now we're here. DSP can do without realative measure.

[@waltersach6679]

How can some of them get there in one try????

[@OttoByOgraffey]

Each mouse is given two runs. The first run the mouse is learning the course from start to finish and storing that information in its memory. The second run is after it had learned from the first run, only without the wrong turns and dead ends, because it corrected those, and or deleted those wrong turns and dead ends from its first run.

[@BayMacDre415]

How the hell did the one with the extendable camera and the computer vision and the insane processing power lose? It went straight to the center of the maze perfectly, first try!!! The others may have been faster but they took multiple tries!! That doesn't count for x4 points or something?

[@jayantabanik2751]

the first try for every participant was to explore the maze ... the participants maybe were allowed multiple tries but the fastest micromouse to the center wins p.s. sorry for foul english

[@RaptorJesus.]

as a westerner i feel like a muggle watching those last few wizards showing everyone how it is done! :P

[@ChristinaKilgore]

Where is Fort Worth Taxes?

[@hood171]

How he identifies that arrived at the destination? Usually at the start and destination there are sensors, but for the time tagging. I could not identify how the micro mouse identifies that arrived at the destination.

[@Greenye]

+Fabio Alexandre Muller there is a sensor at the entering of the center, hiding in the post. I forgot if the center sensor is battery supplied or hidden wiring under the maze.

[@hood171]

+Green ye Thank you for your answer. You know what type of sensor is, and how it communicates with the micromouse to report that he arrived at the destination?

[@Greenye]

+Fabio Alexandre Muller it doesn't communicate with the mouse. it connects to the computer that controlled by the staff to the front. if you want more detail about the sensor, please contact Mr David Otten. You can find his email at APEC micromouse site.

[@hood171]

+Green ye Thank you, I will do that.

[@uncute451]

and i will eat some fish

[@TheRealFreznoBob]

Telescoping camera bot should have been ranked higher but I guess they don't want bots hacking the map by getting a spectators view on their own. I mean what's next, drone air support? OMG that sounds great!

[@shubhampal7229]

where can i get code for this

[@Greenye]

shubham pal u learned and type the code by yourself, bro. Check micromouseusa.com for project futura

[@sathishkumarr7106]

Wow amazing 🤩 so many innovators and inventors here❤️

[@wongmremojeporo7973]

Mantab 👍👍👍

[@hooobi]

is it ever going to move on?

[@deenasmusicbox]

My how things have changed...

[@Hellefleur]

8:58 Can't leave anything behind in the maze. *builds robot that ejects camera out of the maze*

[@vgfxworks]

20:29 was pretty amazing ..

[@TakerAMD]

5:37 is he cheating)?

[@Joshuaxiong2]

7:02 It's like it already know the map.

[@amjadalinava]

Actually it was programmed to have a right turn as the first option in an intersection . Thats why it took the first right and u can confirm it later when another intersection comes.

[@Pe_De]

6:50 👍 pemetaan map

[@Pe_De]

6:41 "WOW"

[@srdjanmalinovic1085]

DAMN Japans one the best with camera XD God damn.

[@FreakShowScitchy]

Why are some of them running twice?

[@NatoBoram]

FreakShowScitchy First they discover the maze, second they solve it and go for the fastest route in the shortest amount of time.

[@repely]

FreakShowScitchy because they memorize the maze first time around then the 2nd one is the fast one

[@Tyunz]

so that's what larry david does in his free time

[@giuseppetello8344]

5:20 AMAZING 17:27 me gusto mas el sonidito

[@Joshuaxiong2]

First place 19:20

[@nigeriansista]

Whats going on? Is it a remote control mouse car

[@DlcEnergy]

1:26 that old man with the mic looks like Larry David (from Curb Your Enthusiasm) and sounds like Obama. edit: 6:50 lol

[@123werty00]

ma che figata pazzesca!!!!

[@IvanRodriguez-em2yr]

como lo hiciste compa

[@surrakasurovec3641]

Why did the dude hit his robot with a friggin rollie?

[@Greenye]

clean to tire to enhance the traction on the following run.

[@JordanMSeverns]

they should just program it to follow the left wall

[@Greenye]

competition at this level usually has its maze design to against wall follower. Besides, wall following is very inefficient.

[@LasVegar]

I tougth what about a robot got high up and tool a picture and boom, there it is 5:15

[@rembrandtes]

for competicion are the mouses bought ?and just programmed ? they seem almost all equal , i do not think they are all home made .

[@Greenye]

they are all custom made, especially at professional level like this..

[@MartinJohnZ]

2:12 NOOOOOOOOOOO.......

[@chuckybraswell2599]

DDOI brought me here.

[@cwv157]

Good ole Fort Worth Taxes

[@royjr16]

NANI KANSEI DORIFTO ?!

[@driverpainter]

MM7 5:28 is Amazing!

[@elmowilcox]

Holy crap the USA is WAY behind Japan in micromouse tech

[@moflyboyblanquito541]

Amazing

[@rodetranx8802]

Fuck with this smart ass engineers..I could even barely connect wires..

[@shrobbes]

I was looking up how to get the free vbucks, how did I end up here? And why did I watch the entire thing? And why do I want to watch more of this?

[@RJI_404]

Thats flawless

[@paulgreenwood4935]

The Japanese guy given 10th place should have been the clear winner on ingenuity alone. Then he should have been made a double winner on safety too. Imagine all those walls that the other 9 crashed into were pedestrians! The Japanese guy's entry is what autonomous cars are all about, ie, learn the route first then travel the route without a single mistake. For example, apply these top ten entrants to a real life road and all the other 9 except the Japamese would have crashed into the first wall and been undrivable from there on. Notice the Japanese guy's car did not make a single mistake whereas every other one was crashing into walls using brute force trial and error. So what they were faster that's not the goal in autonomous driving. The goal is safety and and on that score it was Japanese guy 100%, the rest 0% That fact went right over the heads of the other crews and the organizers and the attending audience.

[@Leviathanshadex]

I noticed the 10th place guy knock into a wall @6:47. Also, this little competition is mouse based. Mice have the interesting capability to complete similar mazes if enticed by something like food. It's this propensity in mice that inspired the competition. Autonomous driving vehicles will rely heavily on GPS, the goal will not be unknown for such vehicles. These mechanical mice clearly have a completely different job than simply relying on GPS, and it would be a fantasy to attempt to translate them into something like the vehicles that you see on the street.

[@rodetranx8802]

Think of it this way imagine that being a spy equipment and they would try to send that thing of on a vent..which means there's something on the top..it can't really picture where it's going..also the only way where he can sensor any obstacle or way it could go is on the beginning imagine all of the sudden it came to a pedestrian lane and peoples are about to cross he didn't really pictured that in the first place..

[@MusculosoDigital]

imagine that robot in the backrooms

[@lastmanstanding3739]

8:08 that had got to be cheating? How the heck did this one know

[@Joshuaxiong2]

Dang, fast.

[@4lifejeph]

I want to see all of these robots in a bot battle for supremacy after this competition

[@Joshuaxiong2]

They should make it where no one knows the design of the map.