Week 1: (8/28)

I attended weekly lectures. This week consisted of me debating what I wanted to do for my project. I'm considering doing a robotic arm as I recently viewed video on youtube of some cool robotic arms.

Week 2: (9/3)

This week I purchased my controller board from PVR. I decided it's not worth the hassle to make my own and this board does most things that I want. In addition to taking more time, I decided in addition it likely wouldn't save me any more money and its also nice to have some pre-written code that I can use to get myself started.

Week 3: (9/10)

This week I decided to chang my project to MARV, a robot who can move on land and in water. MARV will take out enemy installations on land before entering the water to recover cargo.

Week 4: (9/17)

This week I went to hobbytown to look at servos/motors/batteries, etc. I purchased a Traxxas 2056 servo. I was told it worked very well in water and under water. Excited about a new toy, I hacked it open to see it internally and converted it to a continuous rotation servo. It was a rather simple hack that involved using my dremmel to chop of a tab on the gear above the pot. To remove the effects of the pot, I simply hacked off the top of it so the gear above just slides in its spot. It works pretty well and shows no signs of slippage. I adjusted it to be neutral given 2ms pulse and go to full speed at 1.5ms and full speed the other direction at 2.5ms. It was a rather simple hack. In addition I ordered my AVRcan and an additional 8 servos on the internet, 2 of the servos are the larger 2056 variety, and 6 of them are the smaller 2065 variety, both are of the waterproof type.

Week 5: (9/24)

Just received my servos/AVRcam in the mail. I assembled the AVRcam already, but encountered some difficulty. I soldered a header on the wrong side of the board. Due to the nature of the header, I had to remove it and put it on the other side of the board. Parts of the board were messed up during the unsoldering process and I haven't figured out how to fix it yet. Hopefully it will not be a big issue.

After some internet research on tank tread, I've decided not to try and assemble my own platform (at least as far as the tank portion goes. I ordered two cheap RC tanks off the internet and am expecting their arrival by next week.

Week 6: (10/1)

I received my AVRcam last week as mentioned. I still haven't been able to fix it so have ordered a CMUcam1 as a backup. I figure it makes more sense to order an assembled camera to make sure I have no more future problems regarding the camera.....I can't afford a lot of delays over things like that.

In other news, I've received my tanks and have proceeded to gut one of them out. I'm pleasantly surprised to find the interior compartment of the tank is appropriately roomy and should easily fit all the electronics I need, though likely the battery will require an external encasing which could be tricky, but I should be able to accomplish this. Also, seeing that the motors are fully capable of doing all I want/need, I feel there is no reason to replace them, so I'll be needing motor drivers. I've ordered free samples from TI of a suitable motor driver and have put in a request for having my board made with the TAs hopefully next week. I've also purchased a set of transistors to play around with creating my own h-bridge. I've used FET's for switching before and don't think it'd be terribly difficult if I can get properly rated transistors.

In additioni I've visited Lowe's and Home Depot and have been delighted at the extent of watertight possibilities I have. Between PVC pipe's and plexiglass I should be able to construct something that works. I already have a few ideas in the works involving a C-pipe with a window for the camera/gun enclosure.

Week 7: (10/8)

Unfortunately it's more difficult to get something made in IMDL than it was in senior design. I submitted my motor driver board on Monday and on Wednesday afternoon it still hadn't been made. I was assured by Mike he'd make it next week. I've decided to quick order Motor Driver from Sparkfun.com just in case there are issues...I must say I'm dissapointed in lab availability and speed of getting boards made.....dissapointed.

While I was at radioshack purchasing transistors for testing my h-bridge design, I picked up some CdS cells and IR sensors. I'm not sure how I might integrate these in my project, but I think it'd be best to have more than just my CMUcam for navigation. My initial idea is to use IR sensors for line following to go from my "base" to the edge of the water. I can make a solid line/box for MARV as a targetting area. I'm thinking of using a CdS cell to check for light from perhaps a flash light attached to the inside of the base which I turn on the flash light and MARV activates from his hibernation mode.

Week 8: (10/15)

Initially the chips I ordered from TI worked, but something happened and there is a short on the board which I can't find. It's a good thing I ordered the Sparkfun boards. I've set the drivers up and they work beautifully, though they do heat up and after they've been on for a while the power fades a bit, which could be a problem, but I'll have to deal with it for now, I can't afford to spend any more money on this project on the land portion, I still have a lot of stuff to purchase, budget is hitting me hard.

On a good note I got the CMUcam and have also begun some testing on it. It needs good lighting to work properly I've found. I had to use it in a well lit portion of my house to get good results. In low lighting it can't reliably sense color from more than a few feet away, but it can sense from over 10 feet with some reliability in well lit areas. I think with further work I can get better performance and it'll function for what I need. I've seen it work, so it must.

I've constructed a waterproof battery encasing out of plexiglass parts cut out with my dremel, a tool I've found very helpful in this project so far. In addition I've constructed a camera housing mount out of a servo and a piece of plexiglass. I can't find servo horns for my servo here, but I found something called a servo saver that works beautifully for this function. I wish that the Traxxas servo's had better datasheets (or any real datasheet at all) I have to guess at all the dimensions or measure them myself with varying success.

Week 9: (10/22)

I ordered an airsoft gun this week and am expecting its delivery by the weekend.

I made a goofup on planning the front end of my robot, the battery is just too much weight on the front so I've constructed a wheel attachment to go on front and balance the robot.

This week I also started my UART interface software. I've tested it with hyperterminal and it works, except some missed characters on really long strings. I wonder if this is due to the error from having a slightly incorrect baud rate, 117kbps instead of 115kbps. I am considering altering the baud rate, but will decide that after I see how it works with the camera.

In addition I've tested the IR's following a line and it seems to work quite well without the top on the robot. This experiment can be seen on the results page of this site.

I noticed recently that the servos do not spin fast enough for use as boat propellers, I wish I'd thought of this sooner and I wouldn't have wated considerable funding on the waterproof servos, now I'm in a financial pickle, but I will purchase the new motors anyway. I need DC motors with ~200-300 RPMs for use as props, especially in the case of the vertical motioni. I feel silly for not noticing, but I'll persevere.

Week 10: (10/29)

I've been really sick for the past week but have been trying to accomplish some things.

I redesigned the platform for instance. I didn't like how it was performing with the line following. That is to say it was performing pretty crappy. So I redesigned it using the alternate tank I purchased to shift the weight backwards by placing a plexiglass (this time removable) platform above the attachable part (unfortunately I did not take pictures of the old design before I disassembled it to get the battery enclosure off) and placing the battery inside of a PVC pipe with caps on both ends which can be removed. I've realized it is smart to include ways to remove water in case of leaks. If properly made, all of my compartments should be able to take small amounts of water without causing electrical problems by encasing the electrical components inside protective bags to protect from "splashing" effects. With the exception of an all out flood into the compartments this should save my equipment in the end. I'm very happy with how it looks functions so far compared with the original design, its simply more versatile.

Week 11:

I'm considering conversing with the staff about changes to my design as current ideals are not looking very good right now, possible, but not good. Perhaps I can pull back on a couple of my ideas. I've got room to work with though.

My UART design works with the camera. The transmission seems to work correctly as I've noticed no real problems in the functions that I ask the camera to perform. It blinks the debug light, moves the servo, gets the mean values and tracks colors fairly well that I can see. I've noticed that my values for tracking the color red appear to be slightly different than before, perhaps due to different lighting in my house so it appears I may need to incorporate some sort of calibration method on startup on site for demonstrations. Will work on that later.

I've also been working on receiving data from the camera, and it works alright. It works best if I incorporate delays in the transfer. It seems my microcontroller has a hard time processing at full speed. I'm also considering a change of baud rate as I mentioned earlier, but have not yet incorporated this part. Currently I'm working on my camera algorithms for sensor demonstration this week.

I've missed a couple demos recently, partly due to the fact that the TA wasn't in the office when I could make it by and that I was very sick during the last demo week and couldn't make it in. Therefore I really need to make it in for this demonstration....perhaps something can be worked out about my missed demonstrations.