Interfacing

Our 2nd motherboard (triple processor motherboard) has been sent in for manufacturing. Boy was the last hour an adrenelin rush...... but we got it in.... and now fingers crossed hoping that all connections were correctly wired.

We'll be testing signals from RC3's (control team) UARTs just to make sure they transmit and hopefully when our tri processor comes back, we are able to test if a proper transmit and receive can be done between the motherboard and the ATMEGA128 at the base board.

Apart from debugging, there hasn't been much done. Ohh!! David showed us some pretty cool soldering tricks. If a pin from your chip is broken, what would you do?? Panic...?? No.. all you need is a filer and some wires.... The filer to file the casing of the chip till you see the conductor connected to the pin that was broken and then solder a piece of wire on to it... and TaADDDDAAA... it's done.

We gotta start linking the Snappers together. So we'll be whacking in NFS(Network File System) and also Minicom(for debugging) into the Snappers ... somehow...

We'll get there!!!!!!!!!

It's Been A While

It's been a while since we last updated this blog. We've pretty much been doing RC RC and RC.
The single processor motherboard has been put together and now works. However, further testing needs to be done.

We're also rushing to meet the week7 PCB submission. The board we are sending in is the triple processor motherboard and boy is it a routing pain. Good luck Khor (he's doing the routing).

Apart from that, we are still getting software done as well.. I'm kinda working on global path planning and motor interfacing at the same time whereas Shubbhy is bz with deciphering information from the vision system.

We'll i guess that's about all the updates we have for now.

Ps: for those anticipating videos, it will be up sometime soon. Our director/editor has been really bz

Path Planning

Soo... we've been mentioning about path planning quite a bit. LEt me just refresh your memory a little bit. Basically, it's the path the robot is supposed to take after processing the data given to us from the vision system.

At the moment the code is pretty much done up. Just some finalizing and integration left. Integration as in combining it with the existing code we have from last year's work and not dy/dx.

We're now waiting for the camera to arrive so we can start plotting points and taking some measurements to compare robot vision distance with real distance. In other words, what the robot sees, is not actually a linearly related to the actual distance it's supposed to travel. Therefore we're trying to plot a relationship between the two to do the necessary conversion. Boringgggggggg

Well that's about it for now. Episode 4 is on it's way, so for those of you keep up with the blog, keep looking out........
tatta

Semester 2 : Where the fun begins

0x440x4a0x200x720x6f0x630x6b0x73

So , I'm back , we are back , everyone is back for Semester 2 and the work goes on.

The beginning of this semester would mostly be writing code , configuring the Snapper board and also debugging the soon to be manufactured PCB and not to forget the finale
--> Multi Processor broad.

RC1 "spying" on David from the Endeavour Team
hmm always wonder what is he doing alone in the room ...

But now whats on our top of the list in coding would be developing the code for path planning. So what is path planning ? It is basically developing an algorithm which uses all the resources available to the robot (eg : a world map of the field with objects present) and finding a path to a designated target on the field. But time does not permit us to write sophisticated algorithms. Our aim currently would be at least to get the robot moving, so our first priority would be just making the robot moving to a target without obstacles and later with obstacles. The code would be developed incrementally.

Jamie , Head of Department at the labs last week
trying not to push the red button.

As simple as it gets, we will just get the robot to turn to the direction of the target and move towards it. Thats the current aim. The code isn't hard with our godlike coding guru Shubham tapping away on the keyboard. A fair bit of math is involved. For instance , location of the robot on the field (trigonometry & triangulation)? Which coordinates are they with respect to (axes) ? Which direction is the robot currently facing (angles)? Local coordinates or global coordinates ( x-y and graphing equations)? Are there obstacles on route (bang) ?

After crunching through the numbers, action commands and some parameters would be passed to RC3, who deal with the actual physical movement of the robot, to do their magic. This we will see in the next couple of weeks. and till then ... we will meet again :)

01101011011010000110111101110010

Welcome Sem 2

It's been a while since we last updated this blog. It's been a busy break. Actually, i wouldn't call it a break... it's more like just a non teaching period.

I have been working on routing the PCB with Khor during this period. Didn't expect it to take this much time but ohhh well.... at least it's out of the way now. Hopefully the board comes back working.

Complications were mainly with routing the high speed and sensitive signals from the camera link header to the LVDS chip for teh Vision system. David showed us a good trick to route those differential pairs efficiently.. Routing one on top of the other. He also introduced some VERY interesting routing techniques.... It really was ummmm..... something different.....

Now to configure the Snapper. it's going to be a bit tricky but we'll manage. Khor managed to find a lot of info from the web and our C guru SHub is working hard on his codes.....

SEM 2 begins... more hardcore... and more ROBO news to come.

He is BACK !!

So , week 12 is coming to an end. Its time to say our goodbyes and hello to exams !
Before venturing into exam paradise, let me show you a sneak preview to the work of Semester TWO !

Nick giving pressure to Nazira and Karen from RC3

Currently, the first Single Processor Motherboard (housing one Snapper Unit) has been finalised(yeah , "sort of", but its there). We designed it using a software call Altium Designer 6 (formerly Protel) in which a schematic was drawn. Drawing schematic its easy, however when u look at the big picture after , I realised that , "hey i have 50 capacitors on my board , thats cool" but ... i still have to route them MANUALLY ... gosh ...routing nightmare.

Me giving pressure to Shubham

Hmm , ever wonder how small our Snapper Module is ? now lets see ...

See this little beast ? Its THAT small, smaller that the standard name cards. Houses the big beast of them all a PXA255 processor (formerly owned by Intel) , smaller beasts like a Cyclone FPGA (Cyclone III will be out soon), LAN interface, UARTS, USB, SDRAM,MMC , Audio , LCD .. etc etc and the lists goes on.







So, how does our motherboard look like ? Wanna see ? but its confidential. So sorry, can't do that :(

Poor Wen from RC2 got his ahem kicked

But i'm kind ... i'll show you anyway ...

So the white paper on top of the green board (base board of the robot) is our "preview" motherboard and you can see where Snapper would be sitting on it in the near future.

The baseboard would be where RC3 (Ball Control Team) whips up their magic.

Hopefully after the exams, we'll be done facing our routing nightmare and the single snapper board would be out in say ...
middle of July ? yeah i hope.


Nick at it again



So as semester two beckons and exam days lies near, good luck to all Fourth Years in their exams and project. Endeavour guys, the BBQ was good. Thank you !

Special thanks to David for the help all semester long !

We close the curtains on Semester One with our little friend ... HE IS BACK ....






See you in Semester TWO !

DID ANYONE BREAK THE CODE ?

Nearing the end of Sem 1 ....

12 weeks ..of sem 1 are nearing an end ..phew .. 12 weeks passed so fast. Was a pretty eventful day today in the labs. Starting right from the morning , Nick calls me up at 10 saying Iven wants to see the rought draft and then was pressed the panic mode button as the report was not even half way but we finally managed to finish it up before 4 PM and emailed him the report , waiting for the feedback from him ...... The email client "Melbourne University's email client " gave me the most weird warning today while I was emailing the report .. "NULL characters found in your email " whats the crap .. can someone please explain to me ..and Dej ( oh yeah ..dej is David our "cool" supervisor :P ) just replied to me that he recieved several copies of my email .. hehe ..

Also had this BBQ arranged by endeavour group for all the 4th yr EE students . Excellent stuff by them and had a nice time there , seeing how can "some" people eat so much , well realising engineers can do all sort of weird stuff. Also had this photo session in the "robocup" labs ..which we will post up soon.

I just realised we havent written any techical stuff on the blog so far , ooh yeah okk lets use some big words and scare some people with these terms . After the change in the design , that is , RC2 deciding to use the same processor board as us ( we are btw ..RC1 ) we need to write Device drivers in linux. We need device drivers to actually get the data which there Cyclone FPGA will process to the block device and so that it could be an input to the AI and can be processed further. Have absolutely no idea how to start writing them and I wish we also had something like this for our project ..


Dej is going to email Bluewaters ( another character in the RC1 story for 2007 , they are our kind and generous sponsors who gave development boards worth $5000 to us for half the price.. see how nice they are :P)

Well as exams are coming up soon so we will be a bit slow concentrating more on them but will keep posted ..till then ciao..

LAyout Change

There's been a slight change. Currently RC3(Control) that is has 4 Microcontrollers will be reduced down to only one. This means less complexity for them but layout changes for us. However the new layout looks good. And we have more space now. Good to also know we have been doing our project correctly.

Midway report is also out of the way so we can start channelling our attention on routing the PCB once exams are over.

Still PCB-ing

We just settled the issue of the many power supplies which we have to have on our board. Now it's just a 5V and a 3.3V making life easier.
PCB design should be okay.. We don't even need voltage translation. Things seem to be going on fine at the moment.

DS3. Thank God I took DS3. Termination will come back to haunt when we design our multi processor motherboard.. At the moment, however, we are still not sure if termination is necessary for our single processor motherboard. Might have to.

Vision is going to sit on the motherboard as well. RC2(vision team) will be using the our (Motherboard Team) Snapper's FPGA for their image processing. Now we can directly obtain data from the FPGA!

Shubbhy.. good job with the vision codes.

Giving the robots their Vision !!!

I just realised how dumb can a Robot be !! It just understands 2 things they are a "0" or "1" ..imagine a human could just understand zeros or one , how boring can life get then.

These days I am writing code for the vision part of RC1 which involves decoding the data which the camera spits out so thats a stream of 0's and 1's to what a "Robot" can understand.

The robot has to first find out where its located on the field first then , it has to look at the target object and then make a path to attack that target object and then decide which speed to run at the object and then move towards it ..

It may sound a bit too simple but its not that easy when it comes to the actual implementation.

PCBBB

we've gotta spit out our first PCB soon. well good news is most of the schematic components are readily available.

Another piece of goodnews. Coding for local angle(angle between robo and target) is easy to find. Our concern on whether the angle calculated from robo's direct vision(directly off the mirror) was the ACTUAL angle, was solved. thanks Girish. Local angle can be easily calculated. Robots can turn to face the target fast. However, this is all just talk. we shall see how it goes when implemented.. i can't wait.

The Field is in the House

So we set up the field today, everyone was there to carry the green carpet and blue coloured goal from Boydex Labs. So now we have a football field in the special labs (yeah Robocup people are given this Special Lab) !! Took pictures of what the robots see as what Nick posted earlier. Wonder when would the next Special Labs World Cup would be while breaking a few monitors ...

But anyway, people always think engineers are boring. Well, just wanna prove a point that. We ARE TOTALLY DIFFERENT. We special people ( *hints* Robocup people) do a lots of crazy stuff in the labs. You know, its quite stressful in the lab when we are up with some problems with software or hardware.

To relieve it we do ...








Yeah, Moonwalking with our lovely orange ball.




and sometimes ...

**Where did u hide my money ! **

And to Endeavour people ! sorry if we bothered you since your office is just beside us :)
Yeah heard you guys were looking for the guy who keeps singing in the labs ? One hint i can give you is , he wears a yellow shirt almost everyday and is a friend of the Robocup people :) Keep a look out for him !

And to the kids , DO NOT TRY THE STUNTS AT HOME !

till then ... c ya !

BREAK THE CODE

Some Progress

VISION.. This is what robo sees. cool huh... 3D vision.

Today's been good. After our long hrs of trying to figure out how to mount the pendrive on Snapper, D.Jahshan got it working for us in less than a minute. Finally we managed to compile some programs.

Now the next thing would be to work on our PCB and hopefully we can get ROBO running soon. We're still waiting for the schematics and footprints from our sponsors. But they shouldn't take long. They have been quite efficient in responding to our mails.

The Awakening

After the long wait, it has finally awaken. We managed to get connect to the module after Khor's countless efforts of trying to gain access.

It was a big run in a loop thinking we needed to install open embedded in before we could get started. That was really a big road block to our progress.

More to comeeee!!!!!! Tmr's project day. Hope things come along well!

Btw, the picture was kindly taken by RC3's very own Nasira

Lost Lost Lost !!!!

After so many days I am coming back to the blog , well we been planning the stuff all the while and now when our stuff has FINALLY arrived after the Longest wait ever we are trying to do stuff. We were waiting for the hardware component from US.
I am working on the vision part of RC. I have to figure out whether the image RC2 passes me is it a ball or not ,but the question is how will i do it HOW HOW HOW ??

The First Entry

Hey .. So this is the first posting .. the first meeting with David was fun and we are really looking forward to get started ..

Shubham