Finally done the thesis

These just be done now, I'd changed lots stuff. Tomorrow, I'll have to continuous working on the patch and "training" robot. Only have no more than one week to prepare the presentation.

Robot running.

Finally, I got robot work.

There were so many problems since I run the robot at the first time. i.e. Errors on robot's and tracking system's coordinate axes, also on the convention between these two; errors on the matrix translation; errors on the angle calculation. These cause the difficulty in real time interaction. As the robot movement is controlled by the human moving, any error on the data processing would cause robot crash to that person or even out of the control. Errors which are on the robot's side were solved or reduced as much as my friend could by coding on the robot control system. What I did was try the best to reduce the other errors by programming in Max/Msp, like matrix translation, convention of coordinate axes. Even like that, if wanna robot move as precise as I expecting, it still needs lots work. But, as our work goes so far, robot can do the basic action with relevant slow consistence speed and it looks not too bad.

Routes for robot movement were reduced from 5 to 4 (or may be 3). Because the data which is received by robot as command are formed only with the value of X and Y. Thereby, the in situ action can not be excused. However, at the begin, this was considered to be coded as one type of the performable injunctive data to robot.

Those 4 types routes are represented with 4 different commands, there are "Moving around person", " Following", " Shape 'Z' follow"; "Parallel moving". Just as it's named, robot was expected to run like that. Until now, there are 2 of routes have been excused a kind of properly. The reason to say that is due to the recently unsolved errors, although most of errors were reduced by coding from the robot's side and the side of Max/Msp.

The evaluation will take place tomorrow with the testing of this project. The first test was two weeks ago, which was so funny. Robot was a sort of uncontrollable, that was due to the Y axis Y between tracking system and the robot, they develop in the opposite direction actually. Now this problem was solved, also the errors of convention of axis X were reduced again today. (Wohoo, looks lots work done, so far)

Anyway, robot is running, and I can do the following work now.That's cool.

First moving

I'm working so far, the robot has not really moved even once. Last week, my project finally ran the first time, however, there still lots problems have not be really solved.

First testing is to send a command to robot, let it follow a person. The command sent to robot is X.Y value, which is exact same point that a person past through. Testing was not too bad, robot did follow the command and move properly. But robot moved to the opposite direction. The reason for that happen, was due to the robot own X-axis was not as same as the tracking system's.
Also , there's a problem about transformation of scale of abscissa between tracking system and robot system.

In the next week, it will be busy: keep testing; updating thesis; evaluation.

Writing thesis

Today, I read the thesis I've done month ago. I feel bad, as I do not understand some of what I've wrote about, and there are lots bad grammars. Correct, read again, still have some "Chinglish". However I did not feel comfortable with this kind of mistakes, I have to keep writing, get the rest part done as much as I can.

Socket problem finally be solved, new problems came out. Today, I just fond some of the control problems with checking previous work for writing thesis. Oh, dear! My darling god! Look, what I've done before! Although these mistakes I can fix and won't take too long time, I just was not feeling bright, indeed, I pissed myself off. Time is getting less, and I have to look it back to fix those kinds of stupidity. Tomorrow I might have to run all of the project. JJ is already not happy with my progression, and my friend is really busy recently...

Evaluation have to be took place next week. Hope, hope, less problems...

Just finished the sound generators Chapter, tomorrow, I'm going to write about the routes design part which will take some time.

Half test

Time is less and less, although, there is one or two weeks left for evaluation, I still have to write the rest part asap.

Until today, the socket finally working, but the new problem is to split the number and the symbol which is stick with the last number without the space. Last Friday, I tested my half project, the sound generate for human. Funny thing is, tracking system just sent at most 3 number in a group at each time, the receive in PD was fine, but when sent to MSP socket, it only showing the first one or separating the number group into two. I tried different way to figure out the problem, until I unpack the receive data with 4 number boxes. The data shows that the first two number were 0, only the last two shows the data which I want. That's wired! But at least got it work.

This test although have not be able to check all the project, it checked the sound generate is working fine at least, then I only have to check the other part to make sure it works only. Hope will get all of these work soon.

Tidy

Finally in Aug. Can not believe it!

Patch I've done was in mass totally, have to tidy and repair it.

Funny thing is , there are exact the same patch, one was work, another one was not. But this was not a big deal, I fixed it. The big deal is, I really need to get two sockets work well then I will be able to do the following test.

Tuesday, hopefully the patch will be completed.

Working on the rest thesis.

First test is coming

Finally, the sound patches are nearly complete. Today I'm going to do the first test about what I've done so far, except the robot moving part. Though, the robot moving part is the most important one, I will not be able to spend much time on it, even if the previous reading are almost about the robot's routes design.
This project suppose focus on the evaluation, but as far as my studying, it seems about the sound design for movement or environment descriptions. Well, how the final demo looks like, it still have some time do work on it. Hopefully, it will be finished on time.

Come on.

These few days was super busy. Just done the sound patches, and made the socket work.
Sound patches I had repeated three times, cost me too much time. Now I'm really in hurry to get the routes patches done really soon.
Mr. Colllins is back, report about tracking system was hand up few days ago. However this report was not like it suppose be, it is won't be a big deal.
Only three weeks left, still have lots work to do.

Dissertation and socket

Trying to get dissertation down by this weekend.
Trying to programming for routes.
Working on the socket.
Done by the end of the this month, hopefully.

Socket is driving me crazy....

I probably have to keep working dissertation over night again. I deserved.

Next week, another 20 pages for JJ. Collins, which is about the tracking system, all the mathematics, I have some kind of an idea, however, it wont gonna work recently.

Still sanguining about the project.

Busy, bussier, bussss....

I don't remember when is the last time I wrote something down here. "Something" I am working on already more than one month, now, still doing it. I thought my project was made progress, at least had done something, whereas I read more, some "something'' it is, not even closer to the goal.

Every time when I try to write what I have done, I just stuck on it. As I feel I'm in the condition of being confused or disoriented. This is a big project to me and it strikes me that so many things need to study and worth to further study.

Beside the music patch I mentioned for a month, I had tested tracking system and socket. These are not suppose be big deal, but socket does have a /some problems. Neither no feedback, nor reporting error. Today I changed another socket patch, hope they will work OK tomorrow.

I admit that I was not open mind really, until I watched Wall - E. Now I wanna repeat patch again, but I do not think "Mr. Time" would give me more than 24 hours each day.... Dissertation cost me all week to sort out what I've done and needs to be done. I found that what I did study are more about Proxemics, but ignored the sound design. Although it is not a big point to study in this project, it would effect to the evaluation part and the completeness. So I have to continuous studying, adding, fixing.

Time past so fast, project progresses so slowly.

Is that difficulty??

Since starting the project, I'd never passed by the mathematic. MIDI, Time, Timing, Distance, Speed, Routs, Computing formula...
It's good that I did not get entangled in the relationship between sound and emotion. Though it's the study point I've got from sound generate which will be the robot language, it's not the focal one recently. Now, I'm restarting the patched for the sound generate, as I'm not going to use MIDI anymore, instead with the electronic sound which is modified by the "groove~". I do not have much time to spend on the sound. I really have to start exploring the routs, that's the most important part. Try to start testing the whole project asap.

Sound, Movement... Movement, Sound.....

Last two weeks, I just fixed a little of problems, but more confusing about how to control the sound elements effectively. Sometimes I even wanna study the what hell is the relationship between every element of sound expression and the human movements. When I imagine the expression of the movements, faster motion always associated me sharper sound? I think it is same to the most of people. Based on I knew, this kind of sound association is according to the impact of personal emotion, knowledge, situation, and so on. So question is: How can I judge what kind of sound suppose happen with what kind of the movement by less subjective?
My thinking always comes by fitful.

Hurry up

I'd bad few days. No working, no thinking, no friend, no talking, just sitting in the couch all the few days, but for nothing. Time past away.

Today, I'm back to work. Reading, trying to write some (probably just a little bit) papers, and continuous to work on the MAX/MSP -- If I can.

As far as the papers I'd read, basically, they were not very helpful to me. But, "Influence of Pleasant and Unpleasant feelings on Human Gesture Motion" and the "Multi-robot Cooperation for Human-Robot Communication" inspired me thinking of the distance and the angle (face to face) in/between the communicational and the "Nature communication". But, how?

Keep reading

Note: Maybe as the best Robot technology is in Japan, both of these papers are from Japanese.
----------------Well, just guess.

Meeting week 2

Since last week I have to hand up the report every week to Mikael. Today, I suppose to give him the first one. I can not say any excuse about what I have not done yet. Sometimes, I just pist off with a little problems. That's bad, really bad... EN.

This meeting solved almost problems I was confused near all the last week. "Who processing the data for robot"; "How to processing - we considered two ways,but not sure which one is better" and the unquestionable question "How to write the dissertation" . Yeah, meeting suppose be like that, or waste of our time.

For this Friday, dissertation, report for Mikael. Also keep reading for Mr. Collins's "homework" and mine.

Everything will be fine, always.

Stuck

I like MAX/MSP，but sometimes, they just do not like me.

I used "router" to organize the list of inlet, each outlet control the sound generate (subpatch). It seems fine that every thing suppose work, actually, nothing came through! I took days try to fix the unknown problem which suppose be in that subpatch, until it became in mass... how stupid I'm!

Well, sometimes I think they like me too. Last nite, I suddenly had inspiration. Let each outlet directly control the specified object to control the sound element. That's not too bad, at least sound came through and the general function works. Enn... still need time to fix other problems...

Hope Guiseppe will be back soon...

I'm working on...

Paul can not help me as his PHD work, but I'm lucky I've good supporter and partner. Few days ago, he phoned me with excitement about the robot's laser and the camera are working. Now we can using the robot's tracking system and it is better that we get less data transformation between different computers.

For mine, MAX/MSP goes fine recently, I'd some scattered patches for the data receive and send; sound generation. I was going to edit some music and put into "buffer". It is fine for just making nice sound, if considering the sound expression, I changed mind. I'm trying to generate the MIDI note. Each time , a new is triggered. Each parameter (pitch, velocity, duration) can be triggered separately (depend on the given number -- considered like distance, time, location, speed). I think this suppose work fine too. Think about the sound control, I'm using "metro","counter", "route" "pack" "unpack"... but still a little confuse to organize these data (not good math).

About setup the camera, we spent all the afternoon to figure out the better way -- where to knock in nails; how to hung the camera stable; apply permission; borrow ladder... finally, we decide to hang a cord between handrails with pulley system. As far as this plan is, we don't have to apply permission, borrow ladder, drill holes... If the camera is setup, to me, the next steps will be the data processing for the robot (might also sound control).

Meeting

28th. May, Paul,Giuseppe, Alex and me had a meeting. This is the first time we've seen the robot. After we know some about the robot, Paul is going to processing for tracking system. Alex will code the robot and open appropriative proxy for transforming the data between Paul and me. I'm going to use MAX/MSP intend of using PD for the routs explore and the sound, as I'm more familiar with MAX/MSP and there are some problems in PD I could not figure out.
Giuseppe showed me how to make the "netreceive" and the "netsent" patch in MAX after the meeting.
Everything seem going fine. Right now, I think the only problem is the routs explore and the algorithm.

Prepare

I'v email to Paul about the camera set up and system building. Waiting for his reply and talked to my friend about the cooperation.
First of all, we need to see how does motion track works first -- is anything needs about the robot code; division of effective region; environment test; data trasformation... at the same time, I start to build a system for producing and manage music/sound by using PD.
Last week I gave myself a short holiday, so I did not do much work. Today I really start to work:
Choose music/sound; study PD --maybe due to the different vision of PD in Mac, some objects not exist or does not work in my PD; read ralevent issuse in those two books, but I have not found more interesting stuff.

Start

Two weeks ago, I had talked to Mr. Collins for the robot. He give me couple of suggestion and let his student help me -- code robot base on the project needs.
Suggest:
Try to use the camera system which is already in the CSIS buiding -- save time for building a motion track system.
Robot can be coded in Avoidence funtion depends on the requirment of the project -- the backwards for failure of the system that I suppose to build for motion data analysis and the distribution。
After talked to Mikeal, Giuseppe and Paul. We deside that we still working in the original way but will think about his suggestion -- if some parts of system was failed.

I borrowed two books about the Human-Robot Interaction. There're couple of researches about how to give robot personality. These issuses are more ralevent to the human behavior research for making intelligent robot, I was inspriated -- forgive me don't wanna talk about this now as I am not sure that would be succesful, but I will metion it in the later study.