Widgets keep on rolling

An important feature of the Ground Control Station has to be the mission time.So,here is another widget built with elementary programming and mathematics for showing time since mission started.Does this look fine or should something like an analog clock be made.

The Virtual Cockpit

Finally i played a flight simulation game and found out how virtual cockpit works.After getting help from forums and mailing lists,i have been able to make virtual cockpit widget using wxPython(GUI library).And yes,'yaw' has been incorporated too.Attached is an image of it while it was been run on test data.This image shows a small r,p,y reading.If anyone feels it is lacking or if anyone has suggestions,please do comment.

Season II post 1

Season II has begun and we are clear about our approach this time.We have already learnt a lot from the first year and seem to have become more professional this time around.
Updates on Ground Control Station
Java has been replaced by Python and Pythonic way has been keeping me stuck to my laptop for hours.With Python being the ideal RAD(Rapid Application Development)Language,my work has been progressing well.For the previous two months have been at forums/mailing lists trying to figure out which libraries to use and how to use them.Two months into PyGtk and suddenly i come across wxPython which has a really decent GDI(Graphical Display Interface) and have been using it for last few days.
Some work which has been done at the local level:
1.Creation of virtual cockpit(haven't figured out displaying roll there)
2.Creation of a clock/timer showing mission time
3.Reading from serial port and writing to it(and yes this time i don't need to press enter key each time to transmit data)
4.Interpret data coming from serial port via "re" module(beautiful feature in Python)
5.Creating a 3d box to show imu data in VPython

Some territories that i have not yet visited but plan to:
1.UI designing principles(883 pages book)
2.OpenGL/PyOpenGl(have used Vpython,but have been advised to minimise it's usage and completely shift to OpenGl)
3.Mapping:Contrary to what i did last time i am tryinfg to figure out how to use any open source GIS to my benefit,otherwise i have my older approach as a backup anyway)
4.Integrating the data with SQLite instead of simple text files acting as database
5.If needed shift the entire code to wxWidgets(C++) to make it lightningn quick and processor friendly
6.Make an installer for the application for all environments
7.Try and integrate Matplotlib(open source equivalent of Matlab+Labview)for all the analysis purposes

If anyone feels that there is anything more that i shall do.please add comments

Second flight of Icarus with onboard camera

Sig Rascal 110 aka. Icarus had its second flight and the first one with an on-board camera, though it was just a cellphone taped at its bottom but the video gives a nice idea about the stability and degree of vibrations of the aircraft. The aircraft was flying over the college's campus at about a height of 80-100 ft.


http://www.youtube.com/watch?v=V_o8M8SRYGo

AHRS Update

Yes, I know we've been missing for a while but we're back and we promise we haven't been idle.

The Attitude Heading and Reference System(AHRS) is two-thirds ready.
We have successfully implemented a 6-state Extended Kalman filter(EKF) to compute the roll, pitch and yaw.The roll and pitch are working astonishingly well, while the yaw is completely screwed up!!

To test our AHRS, we did an experiment in which it was tested along with a commercial AHRS - Xsens Mti IMU .The two units were aligned along the same axes on a bakelite board and different motions were given.The data was logged and then plotted in Matlab.

The results for roll and pitch were nothing short of spectacular.



As can be seen from the graphs,performance of our AHRS almost exactly matches the MTI
leaving aside the initial offsets(due to errors while assembling on the board).This is a huge step forward as with just the roll and pitch, the control system will be able to fly the plane with stability.

However, the yaw is still a matter of concern but we hope to work it out in the next few days.Hopefully, our next post will be our last for the AHRS team.

A big Thank You !! to the AUV team (Autonomous Underwater Vehicle) who lent us their IMU for testing.

Maiden Flight of ICARUS !!!

Yes, the damn plane can actually fly !!!

The first flight of Icarus ( named after Daedalus' son of the melting wax fame) thankfully did not live up to it's name and was a resounding success.

The MECH team, often accused (behind their backs...so please NO ONE tell them!!) of delaying work and wasting too much time in frivolous stuff, have done some fantastic work over the past 3 days to get the plane ready to fly.

It was an amazing sight to watch...the plane is really stable and flies well, so thats good news for the control team.
Somehow today watching it fly I couldnt help but feel that we are a lot closer to our goal..and with a little more effort we can definitely make it in time for the competition. But anyway, that's still a long way off.

A video of the flight will be posted pretty soon.

Suraj

www.dceuav.com LAUNCHED....!!

Finally....!!!
After all the sending and coming back of Demand Drafts, the website of the project has been released....

Another Brick in the Wall...!!

Cheers....

Visit www.dceuav.com

Ground Station Phase I Over

Phase I of the Ground Station has been completed.The Java Code has gone well over a 1000 lines(we we have a target of 7001-to beat the West Florida Team who used 7000 lines),but thankfully each and every line of the code has been documented thanks to Akshay and Gautam .During this phase following work has been completed:
1.Creation of Uplink:
New flight parameters may be added while the UAV is flying,via communication to the SBC.
Data log of the same has been created.Uplink has been successfully tested locally via serial communication between 2 pc's.

2.Creation of Downlink
Instantaneous flight parameters will be received from the SBC.This sub module like the Uplink has been tested.

3.Current UAV location
An image has been deployed in the same frame as other sub modules and drawing has been done.Here in the image ovals show the waypoints and the rectangle current uav position.

With the local testing done,its time for optimizations and implementation of the Horizon.
Also work on RA's end i.e. a GUI to show his camera images must also get started this Tuesday or Wednesday.
Awaiting comments and suggestions regarding the GUI design regarding choice of colors.

Nipun and Ground Station Team

Ist Ever Sensor Fusion !!!!!

Yes we did it !!!
The pitch of the aircraft has been determined by fusing the pitch rate gyro and x accelero.
Well it was not that easy, it took us some time to understand the basic idea of Kalman, then we made our own program of Kalman (which was a disaster) and finally some wonderful work done by Tramell Hudson of Rotomotion inc came to our rescue.

Thank you Mr. Hudson
watch this space for more exciting news

Next Target: EKF (Extended Kalman Filter) to compute roll of the aircraft.

Chak and Joseph

Current Status

Suddenly the situation is becoming very urgent and critical deadlines are coming closer....
It's kind of like the end of a movie when after a lot of suspense and buildup everything explodes into action in the last 5 minutes leaving u breathless and barely able to keep up.
Now we are entering into April and the targets have been clearly defined:
Sensor fusion: Finish estimation of roll,pitch,yaw before 8th April
Ground Station:Basic work done by 15th April
Math Modelling: Ready by 23rd April
First Flight of Sig: 20th April
If we can achieve these aims,there is a good chance we can make the UAV in time.
Lets hope for the best.

Suraj

Next week = This week -> next

Hey guys,
It's been a pretty long week, tons of things happened....sig finally arrived, components undelivered, presentation postponed, account tending to zero, wap still not installed...but overall just one good thing happened- the sig . Since chak has already dedicated a post to it,so let's get into 'behind the scenes' of the other stories.

Last batch of components ordered using the seed money include the camera,video TX/RX, frame grabber,development terminal ATMega128, Serial Port Dongle Programmer, magnetometer etc. Two orders were placed at Sparkfun for the embedded items, out of which only one has been delivered, that too after a long delay. The other order shall be arriving in 1-2 days. Order placed for the camera at Sify, has been cancelled by them due to non-availability of stock, but I am still waiting for the amount to be refunded back. Frame grabber is on its way to the UAS lab. The main cause of worry for me until this noon was TX/RX order placed at BlackWidow. I say until this noon because after almost 15 mails, Bill Strong took few minutes out of his busy schedule to reply back to me regarding the order status and further details, hence I believe the package will step into the UAS lab hopefully next week.

Last weekend everybody was asked to bring a pair of formals for the big presentation to be given in front of the Technical Committee, Dean IRD and hon'ble Director. But due to lack of communication, it was only the UAS team, plus it's faculty advisers, physically present in the senate hall. The meeting has been hopefully rescheduled for next Tuesday. This time, I am not bringing any formals (I think they are jinxed!!).
To successfully complete the project this year, the ream desperately needs the second instalment of Rs.2.5 Lakh, that too before the next week ends. All the paper work for the same has been almost done.

Internet is one intangible component of the UAV without which I think nothing can be done and it's lack of it in the UAS lab is definitely a huge problem. Wimax was considered as the best solution for this but money constraints and a promised high bandwidth and speed from our CC head led to the acquisition of the Wireless Access Point, which is yet to be installed. Again, I am 'hoping' that it will be done in the first half of the next week.

Now some update on machine vision, I have completed the Phase 1 and started studying neural networks. A feed-forward network designed by Alexander Faaborg is the best I could find till now. I will be starting coding it asap. Moreover, I have mailed Mr. Faaborg, who is currently working on the user interface of Firefox 3, to guide me a bit. His homepage :http://alumni.media.mit.edu/~faaborg/index.htm

That's it for this week, I will be back with more interesting stories next week (which will obviously be longer than this one), till then 'May the force be with Us'...

Rohit Arora aka. RA

SIG is HERE !!!!!

Yes dats true our main vehicle Sig Rascal 110 (with a wing span of about 10 feet!!) is finally here with us and its really HUGE..
After starting the project more than an year ago we finally have our own vehicle..
Things are finally taking shape, we have almost all the components we need and yes the next task is the software design and electronics hardware integration.

Long Road to go !!

Rochak

Macine Vision Plan

Working alone and most of the time at home, the team doesn't get to know what all I am doing and how far I have progressed,and since I have fallen prey to this blogging virus I think its time to post something regarding that.
This is a short report on my progress and the plan I will be following.

I have divided the entire task into three main phases :
Phase 1
Implementation of IP algorithms to segment the image into its component objects.

This was done with the help of the color data and then using their relationship with the other objects in the image. The techniques used for this process are not specific and differ for every problem parameters. After narrowing down to two algorithms to process the color data, 1 of them is working properly and giving satisfying results. The second step was completed with the help of a very popular greyscale segmentation technique.
The last step in this phase is to cut and scale the identified blobs from the image.

Phase 2
Using Neural Networks to identify the target alphanumerics and color

The neural networks will be trained with the input data in the form of a vector,from the last phase. And the output will be printed or sent to the VDU.
Once the nets are working fine in MATLAB, the code will be ported to C/C++ .

Phase 3
Integration of all the hardware : camera,wireless adapters,converters

This will be done as soon as all the products are available in the lab.

Rohit Arora aka. RA

Components here finally !!!

SBC.....IMU..... and by next week RA's camera.. we can finally see things happening in the lab.And inevitably, just as things are getting exciting,the dreaded "even sem mid-sem week" is coming up( in which the odd sem results come out too!i mean c'mon!!!).

Assuming we survive( somehow!),we're in for some really long nights as we shift into top gear cuz the competition deadline is fast approaching.But hell, those are our favourite working conditions anyway, so it's all good. After all, there's nothing like a healthy mix of stressed out idiots screaming at each other, with frequent bakchodi sessions and a LOT of visits to maggi baba to get the technical juices flowing.So i guess some pretty wild times lie ahead...

Anyway,RA seems to be getting all steamed up after the "discussion" last night.Well, good for u dude, makes this blog come alive and be rest assured we wont miss out on any future opportunities to hurl abuses etc on you.

Lastly, as i've given up all hope of the "mech team" ever posting here,i'd like to mention that the SIG along with engine + accessories is almost here and should be in the college in 1-2 weeks maximum.

Promise to post progress on linux and navigation soon

cheers and so long
Suraj

The Camera Issue

After the uber-heated discussion in our lab last night about how the Toshiba PTZ camera is completely obsolete, loads of indegenous abuses were thrown at me. Hence, the dormant blogger in me booted up to life and here is my first sweet post....

The small cameras that the team saw are similar to this one:
http://www.theimagingsource.com/en_US/products/oem-cameras/usb-ccd-color/dfm21bu04/
DFM 21BU04
Dimensions H: 50.6 mm, W: 50.6 mm, L: 50 mm
Mass 265 g
Supply voltage 4.5 to 5.5 VDC
Current consumption approx 500 mA at 5 VDC
So lets take this as an example and proceed...
Some of its drawbacks are average image quality,no inbuilt autofocus(cause images to be very blurry and the objects in the image are thus difficult to extract from background),no light sensitivity detector (clouds,air densities will change the amount of light entering the camera lens), the camera lens has to be manually changed for different focal lengths testing (focal length depends on the working distance,object height and width and CCD format,moreover there are standard lenses, our requirement is 59.96 mm (WD=250ft Object H=10ft CCD type=1/4"),there is a 60 mm lens available on this site but whether it will be sufficient can only be commented on after testing few lenses within this range.)
If the respective departments think the mechanical and electrical requirements are well under limits then I still wouldn't mind working with this camera. Moreover, a better quality digital camera has already been ordered,hence,if the team again decides to throw some unnecessary heavy custom abuses at me ,there will be no hardware constraints in switching back.

And thus I would like to end this post with a humble and happy note by thanking everyone for their kind and ever-ready support.
Thank You

Rohit Arora aka. RA

Ground Station Updates

Well tired of serial programming and trying to figure out how to make it work in linux(already done so in Windows),i was frustrated.For days i have been trying to figure out how to make it work..
But now i guess it might prove to be a blessing in disguise.I have since moved on to TCP/IP and for sure it's damn fast and fortunately i am able to achieve file transmission via client-server methodology.It took me less than a second to copy a 50 MB pdf from one location to other on my system.Hope Mr.Suraj hurries up on his side(transmission)so i could start building on the client side.Enclosed pic shows running of the Java program(seems very soothing to my eyes,after about 200 unsuccessful builds for serial transmission this one finally worked)
And btw Linux is very time consuming.Takes you Hours to figure out things you would not even care about in Windows.

Nipun

Ground Station Updates

Finished with file handling via GUI,also done image manipulation i.e.marking of waypoints on the image..Serial port communication has also been done..That leaves us with a very small task of integrating sub modules into a bigger module and then off we go ready for testing.
Images show marking of waypoints on a given map and the other one shows read write operations in GUI.

Navigation sample pics

The pic with straight lines is the ideal path to be followed
The waypoints are (x,y)

50 100
640 150
560 500
100 580
The paths generated by the algorithm using the 2D math model with transition gain k
k=2 (top image) //since gain is large, motion isnt so smooth due to overshoots
k=1.5
k=1

Suraj

COEFFICIENTS !!!

Well for those of you who have a some idea of AERODYNAMICS would have heard of the terms like Coefficient of lift drag blah blah..
But for those innocent Homo-sapiens who are left yet untouched by this Great field this article may be an eye opener

All the forces and moments in Aircraft dynamics are defined in terms of Coefficients like coefficient of lift due to body i.e fuselage, coefficient of lift due to tail deflection, coefficient of drag due to aileron deflection and soooo on...

If you have to determine the dynamics of your vehicle then you need to identify the forces acting on it. Forces are defined by 12 fundamental non linear differential equations of motion.
In the end for an aerial vehicle to be completely characterized, you basically need all its AERODYNAMIC COEFFICIENTS.
Well these for a simple UAV they are about 50 to 70.(Depending in which frame you are using your equations).

It took Mr. Kaul 2 months to determine all of these and it was Captain Nigel M Jodeh at some airforce base in US , Dr. Robert Nelson and USAF Digital Datcom which rescued us.

So as of today Mr. KAUL is still checking his coefficients

Best of luck to him..

navigation update

Good news everyone. Have tested the algorithm with a simplified 2D math model of the UAV dynamics and the results have exceeded my expectations by far. I am now pretty sure that it will work in the final UAV navigation module. The next thing to do is to test it with the actual UAV math model. However, it seems that the control team has run into some serious problems right now with the mathematical modelling, so it will take some time to sort things out. Therefore, navigation will be on hold for a while and I'll be concentrating on linux for now.Will post soon

Suraj

Embedded Update

As Suraj & Nipun told you about Linux and Ground Station. it's time to know something about the Embedded Work.
From the beginning of this project i and Sahil were working on the PIC18f 4431 microcontroller , well you can include Chak also because he is part of each work and he is always available to give his advice on our work any time. Now i would like to tell you the reason behind selecting PIC18f 4431 for our project. We inherited all the stuffs( microcontrollers, MPLAB ICD2 Programmer, software, ICD2 board, etc) required for working on PIC microcontroller. As UAV team used this microcontroller.
But we didn't have PIC Development Board so we needed to do a lot of 'JUGAAD' for programming the microcontroller. ICD2 Board which we had,was a board with some grounding problem so it was a nightmare to work with this board because sometimes it gave wrong reading. But finally after a great effort we were able to flash LED, generate PWM wave,carry out analog to digital conversion. Although Serial Communication was not successful but we succeed in moving the servo moter at any angle.
After some time we came to know that PIC18f 4431 does not support the serial communication. So we had to change the microcontroller.
Finally we decided to work on ATMEL atmega128 microcontroller. The reason behind the selection of this is that we get a source of help for learning this microcontroller . And now i would like to tell you that it support the serial communication. We are working on this. Currently we have been able to carry out a to d conversion, PWM generation, Serial Communication & SPI.

RAGHVENDRA

GROUND STATION SAMPLE

Still far away from the main design,,some progress has been done on the ground station..here is an image depicting the capabilities of java..a simple program which i just created..click on the image to enlarge.and importantly this image is of a java application and not merely an applet..

NIPUN

Ground Station Updates

There was a time i made fun of all those who joined summer courses for learning java..But fate has the last laugh .So here i am studying java with a couple of other fellows Gautam and Akshay..To begin with java is very simple..its all about classes .the gui part seems a bit difficult..well atleast for gautam's HP laptop to handle..we were not able to exit our small GUI programs..and then came the famous ctrl+alt+del to our rescue..
More progress has followed..Now we are able to add mouse events and images onto our Gui as well as closing of GUI by clicking on buttons..
JAVA will definitely be mastered..and i am pretty sure we will be consuming lots of JAVA(cofee) from the night mess before we master this wonderful creation.

NIPUN

The Curious Case of our Proposals

P.S although I am posting this on Valentines Day but the following post has nothing to do with what an average boy thinks of the word Proposal with reference to Valentines Day. Any similarity is purely coincidental (although that may be the case with many of us out here reading this !!!)

Well to start with we decided to look for a Govt. sponsor for our project like DST, DoD(Dept. of Defence )....
Episode I
We decided to send our proposal to ARDB (Aero R&D Board). With a lot of effort Gautam found the form to be filled. With even greater effort it was filled (with maximum similarity to the form which "The Earlier Team" had filled) and SIGNED by all our Investigators and Advisers...
Then the form was printed and taken to ARDB (address was taken from ARDB's website) and lol!! there was No Such Thing as ARDB at that address. Then through some JUGAAD it was finally delivered to ARDB.
Episode II
The team decides we need a backup as well (what if ARDB rejects us ???). So an advice of Mr Sahil Kapur, the team approaches DST where Mr. Kapur says, we will surely get a spons. A little search on net reveals that DSIR is the most suitable agency for us.
We again start filling the form.. again Investigators and Adviser sign it and the form is sent to DSIR...
Episode III
The team's Communications Expert Mr Bahl now discovers a contact in DRDO. Meanwhile ARDB had already rejected our earlier form on the grounds that it had nothing technical in it. They give us time to resubmit. We fill it again and they send it back saying that now you have given us the technical report but no form was filled with it...
So any ways, Mr Bahl's contact "Our Missile Man" says," you dont submit the form to ARDB but send it to another board ( Name of the board is classified by the team)". He says (Missile Man) "You make me your adviser ( one more adviser...) and I will give you my cover letter and then no one will stop your spons. in DRDO"...
So the team happily asks our Director for permission to submit the report to the board and make the Missile Man our adviser...
Episode IV
The Director spoils the game now. He says you dont submit the project to DRDO as they will classify your work and if there's some problem with your research, they will Court Martial you (yes those were his actual words !!). He said you think over it for 24 hours...
The next day Mr. Kapur urgently needed a letter to be sent to DSIR from the director saying that the UAS is actually a project of this college and blah blah... so now it was decided that we had to get that letter to be sent to DSIR and so, we could no longer say that we wish to get spons. from DRDO etc..
The Director himself drafts and sends the letter with all his BEST Regards and Wishes to DST..
Episode V
Now this took place exactly a week ago ...
Mr Kapur calls me late Sunday Night.. "Be ready to go to DST tomorrow to meet some one". He adds that DSIR cant give us more than 2.5 to 3 lakhs..
So an monday we (in suits for the first time) go to DST with a ppt and proposals and brochures... for sponsorship..
and there we get the next surprise...
"Fill a new form for a new agency".... and then give a presentation to our "PEERS" in 2 weeks (PEERS according to DST are profs from IITs and other aeronautics experts who can judge if this project is feasible for II nd year students of DCE)
Episode VI
We have filled the form, it awaits signatures from our Investigators and Advisers...

CHAK

linux : Updates

Have finished process management.Was a little tricky but pretty easy in the end.Now have an idea how to implement the software framework of the UAV.
Am going to begin with the IPC , interprocess communication.Basically, it involves the mechanisms by which the various processes like the autopilot,navigation,data telemetry etc running onboard the UAV will communicate and co-operate with each other.It is a very important subject and if I screw it up , there will be hell to pay.
Leaving you with that cheery thought to ponder,bye for now, will post soon.

Suraj

Linux Updates

done with low level file I/O using system calls.
and more importantly, today i wrote my first program in which i created a new child process from a parent process.In other words, the first hurdle has been crossed.
Things are shaping up, guys.

Suraj

linux: a little explanation

i guess nobody quite knows what i am doing rite now.when anyone asks, i simply say linux.Well,let me try to explain what i am doing in "linux".

Basically,we require process management,memory management,device management(feels like i'm doing an MBA!!) and system administration in the UAV.So right now am studying these things.
Is tough at times but everyday i get to learn something new.That's one of the reasons why linux is fun, i think. Anyway will keep everyone posted on progress.

Suraj

updates:Navigation

well,

the navigation algo is ready to be tested in simulink.
However, it needs a certain function which can only be provided by the control system
team.So we're kind of stuck right now,waiting for abhay and chak to do their magic.
Plz hurry up guys!!

Suraj

SBC coming closer

Well got the quotation for the sbc today.
Seems that by by next week, u can expect to see me playing around with it in the lab.
can hardly wait!!

Suraj

Finally an account and MONEY

YES THATS true

We finally after 4 months have an account for our project and money in it too.
Although college has given that money but still!!!!!!!!
And yes we have cleared our old bills too!!

YEPEE!!