Goodbye!

Hi all. Finally remembered to sign off on the blog! Project is done report is in and results are already out :-) Hope you enjoyed the show.

Nearly There

Ok im way behind on my blogging so heres what i've been up to. Fixed the problem i was having with the LCD. Its now displaying the cues. From an implementation point of view im pretty much done. I think i might make a few changes to the python code but nothing major. I also want to tidy up the website and the way cues are displayed, again only a few small changes. I'll get this done Saturday. Final draft of report is due Friday so i've a few things to finish this evening. Tomorrow i've to get the final evaluation done and added to the report along with a data flow diagram. Hopefully that will be the final draft done. Wish me luck :-)

Pouch

Ok got the pouch finished today, which somehow took hours. Sewing is really hard. Also soldered the circuit and went to hook up everything but ran into a slight snag. Ok so theres two serial ports on the gumstix, one for the LCD and one for RFID reader. Reader only works on second one for some reason so i had to use the first for the LCD but it turns out when the gumstix boots up it outputs its startup messages on the first port, which looks like crazy code on the screen, so i tried getting them both to work off the second port but no luck. I decided then to just have two different batteries one for reader and gumstix, powering gumstix will just be for demos and stuff if i dont want it plugged in all the time, then second battery powers the screen. Turning on the screen seperatly avoids the crazy output when the gumstix boots. Ok thats fine for now however there seems to be something funny going on with the code now. RFID stuff works but not all messages are being sent to the screen. I cant see whats wrong at all. I'll give my brain a break from it and take another look tomorrow. Will make a short survey like thing before i go to bed about the cues evaluation and get some people to do it tomorrow.

LCD update

Finished off a spot of coding today. The LCD will now display the time and date of when u last had something. It does this in a loop. The screen only gives 32 characters, and doesnt scroll which i initially thought, so the info it displays is fairly simple. Just the item with the date and time of when u last had it. The only thing left to do with this is to solder its circuit properly. I'll do this next week when im making the pouch. Also want to look into getting the LCD backlight working, not that its really important but it looks better. It works for its self testing setting but not when i send data to it. I think i recall seeing a list of codes online somewhere to make it do different things so i'll have a look for this. I finished off my second draft of the report today and sent it to Graeme. For next week i want to do the survey for the readable cues evaluation, make the pouch and test it, and also now that i have a bit extra time left i will take another look into the bluetooth and see if i can get it workin. Wish me luck!

Progress... Finally!!

Ok finally a good day. I started off today writing a C program to open the Gumstix's serial port and send data to the LCD. Was'nt too bad in the end, thanks for the sample code Steve, and seemed to be running fine but nothing was being displayed on screen? I was testing this on fedora 6 first because you cant compile anything on the Gumstix. Anyway i decided to just try the code on the gumstix but it didnt like it one bit. Before fixing this i decided to make sure data was actually getting to the screen. I went over the circuit again and it turned out i had the transistor the wrong way round. The LCD was still being powered so i didnt realise. Anyway i fixed this and tested it by just echoing strings to the serial port. The screens cursor seemed to react so i figured it was getting data now. Just for the craic i thought i'd try python again because it connects to the serial and sends data in just two lines of code, and also you can set the correct connection bit rate which might have had something to do with echo not really working. Well anyway it worked!! Its also super handy that i can use python because i can edit the python script which processes tag reads to also send data to the screen. Im just not sure yet what actual info im gonna display???? :-/

Update

Ok just a quick update on whats happening. The bluetooth work is going on hold for the moment as its proven very difficult to find a stable buildroot with wireless and bluetooth capabilities. I have a backup plan for later if theres time. Started work on the LCD screen. It was pretty straightforward to bulid the circuit for connecting it however programming with it has proven difficult. I initially tried python with pySerial but didnt get any results. Im going to try C next which i was going to work on today however i forgot that the lab computers have security on the serial ports in Linux. It seems serial ports are a security risk in linux and you can use it to access data. Nevermind i can work on it tomorrow at home. This has given me time to work on my report and i should be pretty much finished my second draft today. I also took some time to get some feedback on what i could call my system. This was less productive...
Suggestions so far are:

• The Colinator
• Find your crap if you lose it machine
• Come find with Colin
• Find that bitch
• Colon
• Find me
• Locate -"This already exists :-/"

A big thanks to everyone for their amazing suggestions...

Still going bald

Ok i think everything that could possibly go wrong at this point has gone wrong so its all up hill from here lol. Had some trouble flashing buildroot yesterday. First filesystem didnt work but i think thats because i didnt pick right format for gumstix. Second one was too big, it has to be less than 16mb. Finally have a proper one just need to flash it over. The annoying thing is all these things take hours. Had to leave buildroot running all night. It also didnt help that my laptop decided to stop working half way through all this. It turns out if the root folder is completely full, then linux cant boot properly. On the plus side i know how to flash the filesystem correctly now, its incredibly simple using hyperterminal in windows, which is what i had anyway to begin with. Aghhh! Ok deep breaths... :-) Also i updated uboot so thats much easier for flashing file system now. Also i had time to go over programming with bluetooth and python again and it seems very straightforward. So for today i need a break from this and will concentrate on my report, but am still determined to get the bluetooth part finished by tomorrow or thursaday!

Tearing my hair out :-)

Ok my goal today was to install the new file system on the gumstix. I built the file system ok then ran into problems when it came to flashing it to the Gumstix. The method i was going to use was to flash it from the cf card however it doesn't work. This seems to be due to the version of u-boot installed on the Gumstix. This can be updated but theres lots of warnings that it would seriously mess up the gumstix if it goes wrong so im not doin that. Next option is to flash it over serial. My laptop does'nt have a serial connection so i decided to put linux on my pc. Installed ok but then i could'nt use the monitor because it turns out there is some driver conflict or problem with nvidia graphics cards. Its very hard to use computers without monitors. Grrrr! Anyway what i then did was install a virtual machine in xp and am loading a linux image on it. Of course it doesn't have kermit installed. I feel like i've been here before hmmmm. I've lost the will to live at this point so i'll come back to it :-)

Meeting/ Work schedule

Had a meeting yesterday with Lorcan, Steve, and Graeme to talk about the interim report and the project. We talked about how there will be four drafts for the final report. My first one will go to Lorcan, next two to Steve and Graeme and final draft to Lorcan. I've started working on my first draft and am hoping to have it ready by next Tuesday or Wednesday. We discussed the presentations briefly, which will be worked on in the final week after reports are handed in. I also got a spare wifistix and console-st from Lorcan. Unfortunately there was no spare Gumstix but this shouldn't be a problem. I'm currently working on installing the new file system on the Gumstix but it takes some time so I'll blog later on how it turns out. If this goes well im hoping to have the Bluetooth component finished also by next Tuesday or Wednesday. This will give me plenty of time to work with the LCD screen which im guessing will be the trickiest part overall. Heres my workplan:

• 1st April - 9th April: Bluetooth component and first report draft. If Bluetooth component cant be completed by the 9th I will continue working on it until the 13th.
  
• 10th April - 27th April: LCD component. If I need more time on this I can continue working up until the 4th of May.
• 13th April - 17th April: Second report Draft
• 20th April - 24th April: Third report draft
• 28th April - 8th May: Work on final draft
• 8th May - 15th: Presentation preparation
  

The time periods for completing report drafts are somewhat flexible. My aim is to do at least half an hour work on the report each working day.

Its that time again

Ok its been a while since i've been blogging so here's a big update. So theres roughly bout 5 weeks left to get the project done and i've set myself four main goals. First is to use the Gumstixs bluetooth to keep track of interactions with people based in their phone id and also locations in the form of static bluetooth nodes, more on this later! Next will be to do something clever with a built in screen, need to give this more thought on what exactly it will do which i'll decide once the bluetooth is setup. Also I need to go back to the cues and the website and fix them up with the new data and stuff, maybe add some frills to it if theres time. Last thing is getting the report finished :-)

I've been working on the bluetooth component for the last few days and ran into a few problems. First of all using python it should be quite simple to take in bluetooth ID'd as variables then trnsmit them the same way as the tag reads to the server. Problem is the current
gumstix configuration or filesystem does'nt have certain components i need, which are needed to use bluetooth API in python. After fighting with my laptop a little i have buildroot installed which i'll use to make a new filesystem with all the things i need and send it to the Gumstix. I really didnt want to dabble with buildroot but its actually not so scary now that i've gotten a chance to play with it. I'll get a second gumstix off Lorcan tomorrow so i can do this without making a mess of the Gumstix i already have. Hopefully it will all work well, only problem I can see is setting up the existing code for reading tag data on the new Gumstix so that it still works ok. Wish me luck!

Fun with cues

Took a break from assignment madness today and took a look at the cues to see if i could make them more advanced, or more human readable. Now instead of just printing out a TIMESTAMP it says things like: "You last had your phone just after five" and "you last had your keys at a quarter to seven." Once i realised timestamp is 24 hour clock and not twelve, ooops, it all worked fine. Im happy enough with these for the moment. I think i'll also add something with the date. Instead of just saying "you had whatever on: 12/02/2008" it could say "you last had whatever last tuesday"

Meetings

Hi all. Had a meeting with Lorcan yesterday to discuss the presentations and whats happening next with the project. Lorcan was happy with how the presentation went, I just need to handle me nerves better :-) We also talked about my work plan for the project over the next few months. Im gonna focus more on course work for the moment and do some project stuff at weekends maybe. Theres three main things I want to work on. The cues need to be more advanced so im gonna look at this at the weekend. I'll be looking into using bluetooth which will be able to register people the system comes into contact with and also static blutooth locations. The last and probably most difficult part will be to integrate an LCD screen to the system. Have not decided exactly how this will be used yet, but want to get the other two parts out of the way first.

Presentations 2

We had our second fourth year project presentations yesterday. The idea was for everyone to present their work so far and also to give a demonstration. There was a bigger group this time watching the presentations but they all went really well. After everyone had finished presenting everyone gave practical demos of their work, which was cool taking a look at the other projects. Im happy with how my prototype is so far but i need to start thinking about what i want to do with it next. Got some cool ideas from Steve about using bluetooth so i'll be looking into this aswell. We have an interim report due in a months time so its probably not a bad idea to look at that now if i get a chance

Loose Ends

Went back to my php/mysql code today and fixed up a few more problems I was having. So at this stage with the website you can pick which object you lost and it will tell you when you last had it and what items you did have at that time. So far I've just been testing with three items but I can add more easily and it will work the same.

I've started trying to make the hardware more compact and i've managed to fit the circuit inside the gumstix box. Barely! Lorcan mentioned a possible problem with heat coming from the circuits voltage regulator. I checked the gumstix specs and its good in temperatures between 0 and 85 degrees so it will be fine. I've posted a pic of where its at now.

I must bring everyones attention to a cool little windows program im loving at the moment its called tortoiseSVN and it lets you do all your svn updates and checkouts just by right clicking. Well handy!

I think im pretty much done with my demo for the moment so for the rest of the week Im going to do some of my report and continue editing the wiki.

More Web stuff

Im officially hating php/mysql at the moment. I spent today working on my website. I have a new section where you select what item you have lost then it will tell you when you last had it. This is grand but i also want it to show what other items you had at this time. It doesnt make sense just comparing by datestamp cos this will compare it exactly by second, so im using a select query which will find items which have the same hour and minute as the lost item. The problem comes with setting the hour and minute as being the same as those in the data entry corresponding to the lost item. Confused? Thats ok my brain is mush so im stopping for the day.
Also started my wiki editing which I'll be working on over the week. Heres the link for those interested https://trac.ucd.ie/wiki/GumstixRFID

Update: I've managed to fix this problem. It was annoying me so much I had to go back and work on it. I realised I could just take the timestamp field gotten from the query which returns the last occurance of lost item, and set it as a variable. Then I could just take characters at certain points in the string and set these as new variables corresponding to hour and minute. I can pass these into the query which selects other objects that the user had at this time. The timestamp is always in the same format so its ok to explicitly state what characters are taken from the string. Probably not the most elegant code but hey it works

Webpage

Fixed an annoying problem I was having with mysql today. I now have my php pages setup so that when a tag is read it will check what item the id corresponds to and store that item with the tag id and date/time stamp. I have a basic web page setup, its still very rough, which just shows what you had and at what time. I'll have to add a bit of functionality to it. Instead of just viewing all data it will probably be better to select the item your looking for first. I dont think its necessary to store tag reads continuously so I'll see if i can just have it per minute or something.

Meeting

Just back from todays project meeting. I met with Lorcan and Graeme and gave them a quick demo of the protype so far. We also talked about what I could do next for the discretionary and exceptional project components. This is something I will need to think bout this week. Graeme suggested using two readers in some way. Hmmm. Perhaps a system which has a mobile part and a stationary part. I reader built into a bag or jacket then like a "safe" box left at home. When an item is not at home or in your pocket then it could notify the user that you may have lost something through the lcd screen and maybe with a beep or something? If they have lost something they can log into the website and get some cues. I like this overall idea. I'll give it more thought during the week.

For the rest of this week I will do the following:

• Finish the webpage that will give the user their cues.
• Spend some time on my report
• Get the wiki sorted

Nearly There!

Lorcan sent me his lecture notes on php so I read these today. They were really helpful and
i now understand completely what is going on with the gumstix code. So armed with this knowledge i've managed to fix it up so that when it reads a tag its sending the id to my database correctly. This code will work perfectly for my demo and i can go back to it at a later date to replace or edit it as needed.

Spent some more time today with mysql, just doing some test queries and really just testing the whole system in general. My next goal will be to create a webpage that offers some basic user functionality and human readable ques generated from whats in the database.

Code ugh...

Ok im taking a break from coding to report on my progress. Had another blonde moment today when it turned out hyperterminal is already installed on lab computers. Ooops apologies all round. I spent some time today with mysql and php so i did some editing on my test page, test.php, so it now both inserts into the database and then prints it on screen.

Also started looking at the code for sending the tag reads to my database. Was a bit lost at first but Lorcan explained what was happening and I think im with it now. I've got some of this set up properly I think. The python script runs automatically on the gumstix booting up. When i place a tag next to the reader it runs the test.php page which causes the database insert. Only problem is im not getting the tag ids so im guessing i need to look into get commands. Anyway better get back to it.

More fun with networking

Ok so got the gumstix configured properly today to connect to the router with the correct settings and all. Was'nt too hard but i had trouble finding out what the gumstix mac address was, so i could check it was connecting ok. There is nothing on the net about it, however it turns out that its actually written on the gumstix itself quite clearly. Duh. Anyway blonde moments aside the gumstix automatically connects to the router on booting up.

Ok so I've two main things left to do:

• Get some code going that will send the tag reads to my database properly
• A web site with some basic human readable ques.

Im hoping to have these both finished by the end of this week.

Router Fun

Ok was in work all day so only got a little project stuff done this evening. Spent some time looking into setting up the router. Im fairly sure I know what im at now. I gave my laptop a static IP address then configured the router to enable portforwarding to this address. By connecting to the router it should be possible to access the server by typing the IP address into a browser address bar. I'll borrow a second laptop tomorrow and see if it works.

Quick Update

Ok so went back to the circuit this morning because it was wrecking my head and did a new one that works perfectly first time. Cant figure out what was wrong with the other one but who cares it works now. This bout all I'll get done today because im off to work in a while. :-(

Day 2

Well didnt get as much work as I'd like done today, was a little busy. With some help from Lorcan and after several very minor burns and shocks I have my circuit done again. It definately has power connected ok with lights and such but I will need to test it using kermit to ensure its transmitting tag reads. I tried using a windows version as my laptop doesnt have a serial connection but im not sure how reliable it is and i didnt seem to be getting any results. Hope its just kermit and not the circuit :-( So will wait till comp. sci is open again on monday and use one of their Linux machines. In the meantime I want to get my router sorted. I have it reset and connected to my laptop so I just need to figure out how to configure it for what i want it to do.

Back to work!!

Ok so I started working on my project again today. I got my server fixed up after there was a problem with php. I also installed phpMyAdmin which is handy for the database stuff. I had a problem with setting this up aswell but it turned out i needed to edit a config. file to use a mysql user account. So both these problems are sorted and I have the database and server setup with a php page. I put some sample data in the database and have the php page displaying this data just to test its all working together ok. Also took a look over my circuit again. I had thought it was working perfectly but im not so sure now so im going to redo it just to be certain. What I'd like to get done next is setting up the router with my laptop. This is the one thing that I dont understand completely and how its supposed to fit in to everything??? So its gonna be fun. Wish me luck!!!