In my final year of doing research, I realise that I need do parallel tasks simultaneously.  Name it as writing, coding, analysing and lots more. I started being a parallel ‘machine’ during my third year and it continues for my final year ( which is 4 ). At this moment, the my pc is really busy running 40 x 5 test cases simulation runs and I have just finished analysing all the desired results obtained from my previous simulation runs. I have managed to put the paper together and preparing for the final editing, which is writing the abstract and conclusion. This will be then submitted to Jon and next week, I will start preparing the journal paper for the robustness analysis. It will generally take around 3-4 weeks I guess. In parallel, I am actually trying to understand on programming the e-puck robots. Most of the time I will be trying the code when I cannot really sleep at night. I do have a sleepless night nowadays.  🙂

In writing the thesis, I have to take into consideration on the  spelling, grammar, sentence structure and most importantly is plagiarism. ..so lets see what the tool says about this abstract: ( http://www.dustball.com/cs/plagiarism.checker/)

unfortunately it is not free,  a subscription fees of 8USD need to be paid every month ( I will save this one until I really into my writing later!!)

I also enjoyed using this tool: http://www.spellchecker.net/spellcheck/ . it will roughly check the grammar, sentences and gives a thesaurus if you want to find a more sophisticated words..

alhamdulillah 1 paper has been send to jon for his comment…update the thesis chapter tomorrow and ready for another analysis and also preparing for the journal….

can refer to others as well:



the other good one is: http://www.ghotit.com/dyslexia-online-spell-check.shtml


some code snippet

> There is an interrupt (_LVDInterrupt) for low voltage detection on the dspic, 
> see datasheet "Section 9. Low Voltage Detect (LVD)". Below there is my code 
> segment where I implement the event "on_low_battery", which is fired by this 
> interrupt.
> static void init_lvd(){
>       on_low_battery=add_event();
>       on_low_battery->check_function=always_true;
>       set_action(on_low_battery,lowBattery);
>       //set interrupt prority  ->  1
>       IPC10=(IPC10 & 0xFF1F) + 0x0020;        
>       _LVDIE=0;
>       _LVDEN=1;
>       //set voltage threshold  -> 7 (<3.1 V, see datasheet p. 184)
>       RCON=(RCON & 0xF0FF) +0x0700;
>       while(!_BGST);
>       _LVDIF=0;
>       _LVDIE=1;
> }
> void __attribute__((interrupt, auto_psv)) _LVDInterrupt(void)
> {
>       _LVDIF=0;
>       _LVDEN=0;
>       _LVDIE=0;
>       check(on_low_battery);
> }
> //An example action
> void lowBattery(Event *e,void *arg){
>       e_set_led(9,1);
> }

Programming Epuck – Part 2


  • ALGORITHM with ENERGY and DATA ( tested on real epuck ) – END OCTOBER


Part 2:

  • continue the unresolved issues in part 1
  • install ePic2 : ePic2 is a complete framework to command the e-puck within Matlab ( was suggestet not to use this as we can send the data using UART: need to study this)

    • ENERGY and DATA
    • since our robots have moved to the new lab, we can’t really test it now rather have to wait until mid october once we have moved to the new building.

Part 1:

  • tutorial 1 & 2 : linking and compiling ( generating .hex to be transfered to e-puck robots)
  • tutorial 1 & 2: transfer to e-puck ( yet to be tested upon jon’s approval)
  • swarm beacon : linking and compiling
    • update sound.c
    • question on centroid ( answered)
    • other related issues from the code: epuckcentroid.mcp, .mcs, .mcw and the generated sound values files ( ??? ) – old and irrelevant files. just ignore or comment

Since we cannot really tested with real e-puck yet, listed are the tutorials that need to be completed by next week before updating the algorithm.

  • Tutorial 1: very simple LEDs blinking – compiled
  • Tutorial 2: LEDs blinking (pause with timer)- compiled
  • Tutorial 3: selector reading + LEDs – compiled
  • Tutorial 4: send selector position via Bluetooth (simplified method) – compiled – IMPORTANT :
    • issue: pic30-gcc does not recognize standard C keywords (void, return..etc)…..WHY? – solved
  • Tutorial 5: proximity reading with interrupts + LEDs – compiled
  • Tutorial 6: motor speed controlled depending on front proximity sensor values

Programming Epuck – Part 1 – Solved

Previous issues: the *.o files cannot be read, resulting the following errors:

avoid.o(.text+0×0): In function `_readProximitySensors’:
: multiple definition of `_readProximitySensors’
avoid.o(.text+0×0): first defined here
beacon.o(.text+0×0): In function `_detectBeacon’:

it is basically the linker issues, -pic30 cannot really link with the generated *.o files.

Do refer to tutorial 1 & Swarm taxis files for more.

( The .hex file is generated but I changed the sound.c files…need to check back the changes )

Programming Epuck – Part 1

As mentioned in the previous entry, I was pretty much occupied on learning to program the e-puck. The IDE was successfully installed in my Ubuntu and, am ready to test the first tutorial downloaded from the e-puck website.

After few attempts, it is finally compiled and .hex file is generated. Next week task is to install the program both in Windows and Linux and try to send the .hex file to the e-puck that we have in the lab….isn’t is fun to see the robots moving?

I am now preparing the swarm beacon taxis code for the e-puck. the linking looks fine but, in generating the hex file, it gives the following error:

avoid.o(.text+0x0): In function `_readProximitySensors’:
: multiple definition of `_readProximitySensors’
avoid.o(.text+0x0): first defined here
beacon.o(.text+0x0): In function `_detectBeacon’:

there must be something wrong I guess…hmmmm

Not really sure on what is the issues, all the *.s files have been generated, but the errors are unresolved …need to try again….later


This week task:

  • -packing — moving office and house this month
  • second iteration of conference paper ( should send the first draft to jon)
  • meet jon before he goes to Malaysia for few weeks
  • run more experiments in the lab for robustness analysis
  • continue compiling the swarm beacon taxis

Dell Mini and Dependecy Issues

A bit frustrated when I have difficulties in installing additional tools for the IDE, it has been 3 days!!! and raya is coming. The mood has totally turned down by the Eid, there is not much preparation yet, we are again celebrating it with Friends and Family in Telford this year. They have been calling few times to ask us to come down for the preparation. Applied few days leave from supervisor , thus I just need to STOP all this installation and CELEBRATE..aye?. Few more days left for Ramadhan, I have 6 juzu’ more to go and if Eid is on Thursday, leaving me of not be able to khatam this year….sigh…

I have Googling  and trying , there must be some issues with DELL mini 10 and indeed there is!!!!! ‘There should not be any dependency issues’ – they replied to my email…. but why I cannot install them? Even they cannot appear in the synaptic lists?. I think I have to repackage the i386 deb for lpia ( Dell Mini )…will it works? lets see….













Hmmm, finally it works, the answer is simple repackage it, even force architecture does not really work. Finally managed to install everything and need to try a simple robot program like blinking LED. and will then continue to try the swarm omega codes from the thesis and I hope to add my algorithm in it by November so I have few more first extension months on writing the thesis chapters.

I have lots of deadline this few months.

  • Jon is going to Malaysia end of September ( no meeting…:P)
  • We have 2 papers deadline ( 1 in august and 1 in January – basically from my simulation result)
  • Supervising master students project will start on October ( New office and New robotic lab in the New building)
  • Writing up ( which will be based on the papers that we are preparing )
  • Installing robots IDE for my machine

I will be officially extending my studies starting this October….:) , scared , exited ….all in 1 .