Documentation – Basics

FYP Process 9: The Real Built, long process for making the head.

After a long time of not updating, I’ve started to built the actual head for the robot using EVA foam, which was my first time using it. The reason I am using EVA foam is that it is lightweight yet firm, easy to cut and form into shapes, and the joining of parts only requires contact glue (other glue will suffice but contact glue gives a better stick). Plus, I know that I will be building as and conceptualizing at the same time as I do not have any blueprint, every little details on the robot will be an impromptu decision I make during the building process, so low price of the EVA foam will allow me to work more freely so that I do not have to be afraid of making mistake and paying a huge sum of money on it.

To start it off, I converted the paper Husky head into EVA model with 10mm thick EVA as a base layer.

Planning and cutting the paper piece out to form a template with the consideration of how much EVA foam can bent so that the final product could form (10mm thick EVA foam bend drastically differently from paper so I have to keep this in mind when forming the paper template)

The foam is cut with slanted edge to form an angular joint with the pieces at the side, the curves were made by heating up the EVA foam and bent into place, all connections were made with contact cement to make sure it will last overtime.

the blade need to be sharp consistently for a high quality cut so the blade is sharpened after a few cuts.

Test piece were made and resin was applied to test the result of it and to strengthen the EVA foam to make sure they last as long as possible.

To learn and understand the materials fully, the resin-ed test piece were sprayed with black model paint to understand how they will look like with different variable (how the cut affect the outcome, how thick the resin affect the appearance and so on)

details were then cut and added to the head , all using EVA foam with knife, contact cement and hot air gun.

after the front details were added, I’ve dissected the head to add a magnet system that ease the maintenance in the future if it is required.

After which, eyes were added in to test the appearance of the head, the eyes were specially installed in a way that it will give an optical illusion that it is looking at the user no matter where the user is standing.

And then, more details were added to the front head.

The front were coated with more than 8 coats of resin and sanded which helps the paint to stick onto the head. (this whole process took about 2 weeks as the resin require time to be fully hardened between the layers. plus coating them while prevent dripping and sanding is a labor intensive process)

and then, the sanded head were airbrushed with primer (to help the paint stick to the head, and also help me to see the surface quality like checking of small bumbs)

After a few primer coat and sanding between the coats, the head were sprayed and mask to create a clean and beautiful spray job.

for the back of the head, details were built with laser-cut parts while LED were soldered and installed.

then, like the front, details were drawn and cut out from EVA foam and sticked onto the back of the head with contact cement. the LED on the back were also tested to see which colour combination looks the best.

As same with the front, the back were resin-coated with many layers and then sanded and primed, meanwhile, for the internal structure, speakers were soldered and installed inside the ear with properly placed velcro to help with maintenance in the future.

for the front of head, acrylic pieces were cut and melted with hot air gun to provide a protective cover for the eye, which two layer of tinted coat were sticked on it to darken the overall feeling and gives it a more compelling eye to the dog head.(and it affect greatly on the quality of the eye on camera, which I hope people will be taking photo/video of it during the FYP show.)

afterwhich, the dog head were masked and sprayed in many layers with model paint.

towards the final after the base colours were masked and sprayed, the corners were touched up with small brush. Water decals from toy models were added to the head to give it a more interesting finishing and make the finishing completed. As in my research, I found that small details is the key factor that differentiate a normal artwork and a insanely impressive one.

For the appearance of the head, lastly, it was sealed off with two thick coat of samurai lacquer spray paint and then airbrushed with a model grade matte clear coat to remove the shine from the lacquer paint (which is too glossy and look like it is a plastic toy.) small details like carbon-fiber vinyl were sticked to the side of the head to finish it off and gives different texture to the dog head.

Carbon-fiber vinyl after airbrushing the matt clear coat

The process of making the head from start to finish was long, but I’ve learnt many skills as this was my first time using EVA foam, so everything was a good lesson, even from the small details like “how to cut EVA foam properly” is an invaluable lesson for me. And throughout this process, I also learnt that what materials could or could not go with the other, and also how to cause the paint to stick onto the resin firmly and not peel off.

FYP Process 7: Development and Upgrades.

Processor protecting case:

3D Print + Resin treatment as I want to see the effect of applying Resin to 3D printed parts, It was airbrushed with gold paint afterwards to test the paint holding properties of resin, which to my surprise, is good!

after attaching it onto the base of the Robot, this case serve as a protection of accidental dislodge of the pins (while still maintaining the function to switch out the board) from the arduino as well as adding a slight water resistance to it,(you never know if the visitor will drop a cup of water onto my robot during the show)

Parts Strengthening to make sure nothing come loose:

original connection of the aluminium profile. 4 screws with spring washer at each connections on the 90 degree joint.

added a T plate to secure the joint, total screws with spring washer = 8, and it is structurally stable (the screws hold all axis of vibrations.)

Experimental Head for the robot:

3D Paper Model, cut them and stick it together

Dog head. Why a dog? because dog is human’s best friend and you’ve heard of search dog, guide dog, which is exactly what my robot is about, and it is cute.

after cutting the eye out and replace with the LED matrix with arduino nano. i would say it made it really interesting to look at just with the eye that moves.

Plastic model building from the Japan Research Trip:

this is the link of the japan research trip.

The time when I went to Japan, I bought a few of these robot plastic model to study and get inspirations from, so here is it!

Will paint them and display as a diorama at my booth during the show.

Lattepanda with Unity and Arduino:

To experiment with speed and delay of Lattepanda with unity to control arduino, which is not bad!

Lattepanda is known for getting hot, so two 12V fan was added to cooldown the chip of both side to ensure smooth running for a long time.

FYP Process 6: Robot Appearance Researches

show researched :

These are the show I watched in the past 2 weeks to get ideas for the robot personality, appearance concepts, and the overall world for O.U.R.S..

next gen:

This is the nearest concept to what I have in mind, from the robot characteristic, appearance, to the overall concept. the Lab scene, the destroyed appearance is what i could reference on and I really like it.

Final space:

The personality of KVN(Kevin, the round robot) of this show got some really interesting personality, it is really irritating and have some sense of self realization that it is easily replaceable was really cool.

altered carbon:

watched to get a general sense of a cyberpunk theme(which is the world O.U.R.S was placed in) really cool technological stuffs here and the city was really well made.

dragon pilot:

Dragon + Robot, quite cool, except for the fact that the anime was good, there was nothing much for me to reference here.

Wall E:

classic, robotic sound can create emotions too, simple movement and eye from the robot could say alot.

Real Steel:

it is just an action movie with robot, nothing much here.

current Art Direction for explorer 27:

Friendly, empathy, not industry looking, not intimidating.

BY 26 OCT :

unity + touchscreen control master arduino

read book (I robot – Isaac Asimov)

By 9th nov:

drawing for concept

lidar(maybe)

head model (maybe)

Further Research Done:

“Mobile Robot Planning to Seek Help with Spatially-Situated Tasks
Stephanie Rosenthal and Manuela Veloso
Computer Science Department
Carnegie Mellon University
Pittsburgh, PA 15213”

FYP Process 5: going against the Law? Hardware+Software

According to Isaac Asimov ,

the three laws of Robotics:
1) A robot may not injure a human being or, through inaction, allow a human being to come to harm.
2) A robot must obey orders given to it by human beings except where such orders would conflict with the First Law.
3) A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

Sounds really awesome isnt it, of course, there are situations that the robot will be forced to break these law, but again…. what if a robot is made to go against the first 2 law?

1) A robot may lightly injure a human when required to get the task done.
2) A robot may not obey all orders given to it by human beings.
3) A robot must protect its own existence and continue to execute the first and second law.

sounds really funny to me, doomsday robot? AHAHHAHA! sure, why not? ok maybe not. These laws are still in conceptualizing phase and if it made into the final project, it will be quite funny for those who read or knows the three law of robotics.

What was done until now after the last update?

re-structure robot:

As of the previous update(the one before the presentation) whereby I’ve construct the wheel and the base structure for the robot, I’ve noticed that the parts get looser and looser even when I tighten all nut and bolt in every connections, that was because of the vibration when i move the structure around. To counter this problem, I’ve dissemble the structure and change all bolt and nuts to self-tighten nuts (Nyloc nut) paired with a spring washer to reduce the rate of it getting loose overtime. As a loose part will get looser overtime through vibration, if this was not done at the early phase(now), it will be a real hassle at the end as I want the robot to have a relative fast parts replacement (so no joints should be joined with permanent solution like super glue) and this solution of the combination of spring washer and Nyloc nut is the most effective solution for its cost(time and money).

Nyloc Nut, to “lock ” the nut prevent slippage of the nut:

Spring washer, to prevent loosening through vibration

LED matrix animation:

The LED Matrix used is 4 x 1088 8 by 8 LED matrix which will be used as the animation for the eye of the robot, it will be attached to an arduino Nano which will be labeled as “Emotion Processing Unit” in this project to give it a more developed feeling to bring it away from the “Arduino = prototype” perception.

Parola Library code: Robot eye library code:animation made through a program that convert animation to LED matrix codes:

Mecanum wheel coding:

The Mecanum wheel is controlled through a 5 wire motor(VCC,GND,PWM,Dir and FG wire) so the motor driver was not required and I could wire them directly to an arduino. all the code was stored and run through an Arduino Uno whereby the main switch to control the motors were linked to a relay for the motor circuit. As mecanum wheel move unlike a normal wheel, special codes were used to move the robot in any direction without changing the orientation.

Master-Slave Arduino System

There is a function of Arduino that is called the master and slave relationship, whereby a master arduino could communicate with other arduinos(up to 52) through just two analog pin (A4 and A5), this is really helpful for my robot as there was the saying of “Do not put all the egg in a basket” and this function will ease my repair if any parts gets faulty during the FYP show, I could just switch out the problematic parts without re-wiring the WHOLE 5 bazillion wire connections. Also, Arduino is a single thread micro-controller(it can only do one task at a single moment) so by splitting up different processing requirement into different arduino, I could run multiple thread without any lags between them. Finally for the upside of a Master-slave system, it is simply to solve the problem of hardware problem of the limited digital pins each arduino have(more arduino = more pins I could utilize without having more parts/coding like using a I2C Address chip.

wiring diagram

Things are getting technical here…
The Wiring Diagram, as messy as it looks but in normal language, a Master Arduino is controlled by a clickable analog stick (mode 1 to move, mode 2 to rotate), the data is processed and sent over through Analog Pin 4 & 5 to the MPU and EPU which “Movement Processing Unit(MPU)” that controls all the movements of the motor and “Emotion Processing Unit(EPU) controls the eye Animation on the LED Matrix. All of the Arduinos and motors will be paralleled connected to the 12V battery supply to power them up.

O.U.R.S Chief Engineer at work:

Sample Movement of Explorer 27:

FYP Presentation 1 OURS, Explorer 27 the lost robot

SLIDES and VIDEO compilation

The Slides were made with multiple programs,

Adobe After Effect for the animations(I got my animation template online and edited it.)
Adobe Premiere Pro for converting the animation into GIF
Powerpoint slides for the slides.

Why use GIF for animation? Because a video in slides is not loopable and have too much constrains, so if I separate one animation into 2 GIF (Appearing and Looping) and use the “Animation- Appear” function in Powerpoint in a well-timed manner, I could create an illusion that it is looping.

Example:

This is an “Appearing Animation” that plays automatically because it is a GIF, and a “Appearing Animation” will keep “Appearing” because GIF is in a loop, so when stacked with “Looping Animation”…..

This is the “Looping Animation” that automatically appears (With PPT’s Animation”Appear”) 3.2 seconds after the slide starts, so it will look like the animation flow somewhat nicely.

and when put together in PPT, it will look like this, this GIF have the “Looping Animation” played 8 times, after 8 times it will “Appear” again but just an example~ in the ppt, it will loop indefinitely until I click to the next slide.

That was how the slide was made! and I learn After effect for it because most of the animation i used in this slide could be applied in the UI of the robot (Remember i have a screen that needs User Interface, yes, thats the one.)

Also, The Lab coat i wore during the presentation was customized with the logo of OURS to have a more cohesive feel to the theme.

What had i learn in the presentation:

Antropomorphism in robotics

main issue in the fyp:
Fake Ai -> Belivability -> Lost -> Antropomorphism

LEM Solaris (movie and book, george clooney) , I robot russian Stalker

FYP Process 4: Construction, Unity and Lidar

Aluminium profile measuring and cutting

20mm*20mm aluminium profile was used to construct the internal frame because it is easy to work with, lightweight and durable.

(Just for the record, I did not work in the dark but turned off the light for documentation purposes only.)

3D Modeling

This is the mounting adapter created in tinkercad to match the steel bearing, the motor and the wheel to the 20mm*20mm aluminium profile. the adapter for the suspension was created later as the suspension had not reach at the time of modeling the mounting adapter, however the 3 part fit nicely with just one try!

3d printing

Frame constructions and 3D printing part assembling.

many different screws, bolt and nuts were used to construct the frame and attach the wheel mount on the aliminium profiles.

Suspension Addition

The suspension is required as the Mecanum Wheel require friction to move in any direction, so the suspension does not only reduce the vibration, but to force the wheel to have traction with the floor at any point of time even on uneven surfaces.

Unity:
Animation done in Unity,

IlSample quest screen (there will be another screen to show the map or something)

These 3 were drawn in Adobe illustrator in separate layers and imported into unity to animate it.

Lidar:

Lidar uses infared laser to sense the distance in a 360 degree manner, it could give thousands of readings per second and the program will calculate them and map it out through a point cloud.

I bought a Lidar from China to try it out and the 1 I bought was C0602

when using the program that was provided, the lidar works perfectly, but since I am going to use it in Unity, there are problems as Unity runs on C#(a programming language) while the SDK(Software Development Kit) runs in C++, although it is not impossible to write a bridge between the 2 language or write another code in C# to use the Lidar directly in unity, it was out of my capabilities to do so after trying for a week, and while researching for how to do it, i came across another Lidar that provided the Unity codes, however the code does not work on the current lidar I have……

SO, simple solution to this problem, either i ditch the idea of using a lidar OR I buy the Lidar with the Unity code and I put the current Lidar I have for sale online (at higher price than I bought it of course.) And of course, I chosed the 2nd option, to sell the one I have and buy the other 1 that provide a unity code. Carousell, my best friend.

FYP Soul – Why?

Why robot? Why Guiding robot? Why a whole system including a company, a backstory of how the Robot came to FYP exhibition?

First thing first, Why am I building a robot for FYP?

The answer is much more than just because I like it (and of course I do!)
Culturally, there are two opposing opinion of a robot – A western one which threaten us by stealing our and eventually bring us to annihilation, A Japan one which is seemed as hero and seemed to enhance the quality of life, since 16th century after the invention of karakuri puppet, the Japanese enjoys seeing something moving automatically and it is still really fascinating to see something that moves by itself now as we anthropomorphize the object unconsciously. I personally think that robotic will be the next advancement to the world as our computation power increase exponentially, the only physical way we could bring these newer technology into a good use is through something that uses technology and have a physical/tangible characteristic, just like a robot, albeit the term robot was loosely used, the general idea is similar- Physical object that moves without human through a set of pre-determined protocol.

so, why specifically a guiding robot?

This is because I want to be of some use to our FYP batch, Guiding robot’s main purpose is to serve just one function- to bring the visitor to a student booth, which will increase the exposure of the student. Even if throughout the whole show duration, my robot only managed to bring one visitor to one student’s booth and the visitor enjoyed the booth, I would consider my FYP a success as I helped someone(visitor or student) to experience the FYP show in a slightly better way.

How about a lost guiding robot?

For now, I will be building a lost guiding robot which need the visitor to help to locate the student’s booth, although it seemed counter-intuitive to make a LOST Guiding robot as the worst thing that a guiding robot could do is to get lost, however when I go the opposite way(metaphorical), the end result still serve the same function, a robot which guide(narratively, it will be guided) the visitor to the booth. This way, the user experience/interaction with the robot will be different as they will feel like they will be helping the poor robot to find and complete it’s task and the visitor will feel like they have a sense of duty/accomplishment when finding the booth.

How does this work?
All of these stems from the word “Altruism“- the belief in or practice of disintegrate and selfless concern for the well-being of others. In this case, sacrificing the user’s own time to help a random robot.(which by logical thought, they do not need help and does not have feeling, however human is a complex thing and probably will not do things by logic)
As helping others will give us a sense of purpose and satisfaction, I will want to instill this idea into my project to make the user to feel like they are really helping the robot and feel the satisfaction when they complete the task(which in turn makes a happier visitor and a memorable experience for them.)

Why a whole system including a company, a backstory of how the Robot came to FYP exhibition?

This is to adapt the power of fictional narrative to change people’s attitude towards social change(robot in FYP exhibition) by using the method of narrative persuasion- a method that uses narrative transport to persuade us to change our mind, behavior and see the world differently and to put things into context even when the story is a fantastical.

Research to be done:
Interaction of human and robot
Social Robot
Programmed behavior
Slot machine reward system

https://topdocumentaryfilms.com/human-robot/

https://topdocumentaryfilms.com/inhuman-kind/

https://topdocumentaryfilms.com/robot-revolution/

Interactive spaces: Light and Darkness Ver2 Analog + Digital

The final Video:

During the process of adding digital to the analog version, there were more failure than success done throughout, let me start with these failures (and additional works I did which were not used in the end)

The unused Animations

Right after the completion of the analog version, I thought of making a projection of animation from the back of the candles onto a sheet of translucent paper sticked behind the candle shelf, I tried to learn how to make an animation and so I asked my animation friends what program to use to do a simple animation and they suggested Autodesk Sketchbook, so I downloaded it.

I started to draw the head supposed to be shine behind a candle that was unlit.

And I manage to make it move, although simple, this took around 7hours to make sinply it was the first time.

and this is the character that was supposed to be shined behind a lit candle, this is much faster, only took like 1 hour.

The Circuit that works,(or didn’t)

And then the Fairy light I bought from china arrived! I bought 220V fairylight because they are cheaper, and I did not expect them to be these problematic to deal with due to the dimming issue and also 220V = risky.

The light light up really well! they are 10M each and I’ve bought 6 rolls.

and rewire them to a Solid State Relay, I thought this would work really well as they are relays that dont have mechanical parts, so in theory, they could handle PWM for dimming….

and I tried adding PWM code to run the LED, It did not work as there was no dimming, all it did was full flash at faster rate. SO IN CONCLUSION, SOLID STATE RELAY IS FAST, BUT NOT FAST ENOUGH FOR PWM!

Since the SSR alone doesnt work, I thought of adding a capacitor to the circuit, as it will hold and release charge, it should smoothen the flashing rate? thats what I thought. In reality, the flashing of the LED is still really obvious so I cant use it too.

I also modeled and printed a case for the SSR as they are 220V, my common sense tells me to cover all exposed 220V metal parts as much as I could.

As the SSR and capacitor combo doesnt work, I thought of using a Mosfet to control the PWM of the 220V circuit… and my arduino blew off instantly after i turn on the 220V switch. yeah, maybe I wired them wrong, but I am not going to try it again as it instant killed my Arduino… one Arduino down, a few left to go….

So I wasnt expecting this to work, I did alot of research about using the super small resistor I have on a 220V circuit, and everything on the web said that I should not use it, but I thought I might as well try to see if it blows up or get heated up really quickly, at most i lose one resistor, I could definitely afford that! so… WOW TO MY SURPRISE, IT WORKED! THE RESISTOR DOESNT HEAT UP!

AND 1M OHM RESISTOR MAKES IT DIMMER!!! SO YEP. IT WORKED LIKE THIS IN THE END. all those expensive components I bought couldn’t be compared to these small resistors… ALL HAIL RESISTORSSS

so I researched about how to calculate the resistance of the resistor, the more parallel there are, the lower resistance = brighter, so I could do a combination of brightness from a 2M( 2*1M in series), 1.5M(1M and 500k in series), 1M, 680K, 91K. With these numbers, i could do a huge variety of resistance to dim the light(like for example, 2M and 1.5M in parallel gives a net resistance of 857k) , i just need to connect these to the SSR(normal relay could work but since i have these already) and select the path to open to create the parallel circuit and hence controlling the resistance to the 220V circuit.

I’ve notice that there are 5 wires in each strip of LED, and only 3 is requires, and I also dont need to chain these 10M rolls anyway, so I decided to savage these wires in case I needed them afterwards.

and although it seemed simple, it took some time as there are a total of 60m.

I am really glad that i savaged these wires, as just nice Makoto needed this so all time savaging was really well spent!

The original adapter and controller have some problem with it as it will automatically switch mode, sometimes to flicker and such, which I just want to have constant brightness controlled through my arduino, so I decided to remove the chip within it by replacing this circuit.

so I tried to understand how this works and mapped it out

and then I re-soldered it to the circuit board

and in the end, this one doesnt work so I found 1 adapter for this kind of LED light and use that instead, which works!

The additions to my “Dark Room”

I’ve picked up this from near my house and thought that I could build a legit door out of it, so I did, and in the end i removed the wooden piece I screwed in and made it a folding door, I also added magnet to the door so that it will stay closed to reduce light leak into the room.

attaching the LED dtrips to the roof of my “Dark Room” is a big problem as the roof is too big for me to puncture a zip tie through it (i can poke one end through, but can’t poke it back to complete a loop), and tape is too not reliable for my project as i need to have it on for atleast a week, so I’ve use a lighter to burn a needle and bend it slowly to a U shape so that I could pierce it and bring it back to sew the LED strip onto the roof, and just the sewing of the LED took 2 days to complete, because I have to carefully organize where I am attaching each point so that it does not go into the line of sight of the camera I will use to sense the candles.

slowly but surely, I attached these LEDS to the roof and wall of the “Dark Room”

and I tried lighting up, the best thing about suspending the LEDs using a thread is that not only it will create a better illusion of stars, it will also prevent a fire hazard as there is a much lesser chance that any wire(220V) will touch the cardboard and setup a fire. so I am sure of the safety of this project.

Coding is a nightmare

as there are 297 candles, the coding to split each candle was simple but tedious, I am sure there are better ways to do thing, but the downside of using max msp was that the exact function to do things the better ways is really difficult to find, so… my mindset was “If I only know “IF”, I can also do unlimited amount of task, “IF” is as powerful as Hercules.” so… I had “IFs” my way through this project, literally. (just to be clear, I tried to find a better ways, and though I found them, but in the end it either doesn’t work or crashed my max msp.)

the main patch consist of Jit Grab to get information from the webcam, a lens distortion corrector as I am using a 150 degree webcam, a crop to crop away all details which i dont need and to pixilate the average brightness within a certain region to a bigger area(to see is the camera capture a bright object within a certain area which can be used to see if the candles were lighted up indivitually or not.

the jit.cell which shows the numbers in every squares (33*9), which is the amount of candles I have

HEPHAESTUS SYSTEM FYP Presentation slides + speaker notes

Google Drive Download Link

Hephaestus system is a set of 3 roaming robots and a booth, these robots will be specially designed with different appearance and function to provide addition showcase platform and information of all the FYP students (except for me), visitors will not get any information of me from these robots. Only the booth will have information of me because, really, I have to find some place to put my name card. And it is kind of cool for the visitor to realize by themselves that the robots is a work of a student. Another reason for the robots not to show my work is to prevent conflict with my peers, since my robot will be walking in other major’s “territory”, it is simply not nice to broadcast my own project.

Imagine one day when you are lying on your bed and receive a SMS telling you that your FYP project is awesome? how great is that?

Imagine what can the system do for product design student if there is holographic display on one of the robot? You will see the student’s information on the main screen and the student’s work in hologram?

Slightly more detailed drawing of the one that I like the most out of the 36. this will be the rough gauge of the size of the robot

The final robot will probably not be this, but this is a good gauge to show the overall style of the robot which I will be making, not those typical looking or DIY homemade robots.

show the remote control proof of concept i made.

Computing done in the booth system as it will be connected to power supply and heat will be dissipated more efficiently than in the robots. Current location of the robot and movement control will be sent from the booth, user input and message will be sent from the robots to the booth.

Intended recipient and message from user will be transferred from the robot system to the booth through wireless (Bluetooth or Wifi), Booth receive the message, encode it with the recipient’s phone number into the GSM module and send the message through a prepaid simcard.

Skills Acquiring and Prototype R1 goes hand in hand, I will learn while doing and do while learning.

As of now, there are many skills which is essential for me to achieve this project but I don’t know, so I will be learning them. Blender is a 3D modeling software, it is really powerful and the best part is that it is free, the downside is the learning curve is really steep.

FYP part 5- Hephaestus Systems Planning + Software learning + Modelling + Gantt Chart

To understand what to plan for, I would need to understand the nature of the project. As the task for completing a physical + mechanical project differed far from a virtual-screen-based project like games and visuals. There will be more restrictions in doing a physical project than virtual one due to the nature law of physic, material and cost.

Money Problems:

As for building a few robots will cost me quite some money, budgeting will be even more important than time planning, as for where the monies come from, I will probably save up from selling things online and work studies and treat it like a commitment because I think that no one is forcing me to do anything and its all my resolution to fund my own project.
I had thought about asking for sponsorship and that may even happen if I have to. (especially for the batteries that I will be using in the robots, these little things must be of great quality due to safety issues while a good + durable + high capacity + low weight battery cost about $500 and up each and I would need at least 3(excluding spares), which I am totally unable to afford.)

During this few weeks, I had been learning Blender (3D modeling Software) from scratch, it is really difficult to pick up, but i think the potential of Blender is far beyond what I need, so I will stick to learning this super useful program.

I’ve tried to follow a few tutorial and learn the basic of Blender from youtube, this was my first blender experience in building a 3D model.

I stated to learn by building chest as it got similar shape of what I want to produce, and after this, I used the skills i learnt from here and applied into the attempt of my R1 Robot.

the overall shape of this is rather similar to the chest so it took a while to get used to this, however after building this, I realized that I don’t know how to make the top of the robot. So I progressed into another tutorial.

and this was the over shape that I made and I am pleased with it for a first timer effort, although it took me 2 days to reach here, after this, I continued to build the details at the side and front…

Side view

Front view

and I decided to make it like a production poster so I rendered another isometric view to make it to looks legit for my presentation.

In blender, there is a animation function and I thought it would be really cool if I can learn it, so I went ahead to learn it from an online tutorial and produced this.

After I feel that this is good enough for the presentation, I tried to 3D print the model out, it was then then I realized that my model was full of mistakes and it only look good, in actual fact, the surface of the robot was really badly made. So the effort I made in this 3D model got to the furthest here, I will definitely be modeling everything again for the actual robots that I will be building for this FYP as this model doesn’t work, however this was a good learning experience for me and understand that I need to build the model’s surface properly.

the surface detail was not able to be printed due to the mistake I made during modeling, which creates an non solid surface and therefore not printable.

and then it was attached to a small remote control car for proof of concept to be used during the presentation.

now, the Crowd Favourite…..
Mr Gantt Chart!

I started the Gantt chart at 1st of April 2018 as it was all research done till this point.

Within each task, there will be multiple small task which falls into the same category and I will explain them as well as a short description of what it is about here in this post.

Since my project will be physical+mechanical+ technological+ I need to get student’s FYP work early, it is really important to start the execution really early and throughout the holiday because building of the actual robot and troubleshooting the system will take quite some time, and I have to ask all FYP student to submit their work to me really early to make everything work.

Research 1 (1st -30th of April):
I think this is the most important factor of the project, good research done here will help me to reduce work greatly in the future.
Researches up till this point (16th April) – Similar Existing Products, potential parts, platforms for interaction, things I have to learn, Inspiration of artistic works, Parts price comparison, Target market and segmentation,

Skills Acquiring ( 10th April – 8th July):
There are many lacking knowledge and skills required for me to complete this project, like the list of things to learn, I need to learn a few of them to make sure that my system could work. Also, I need to pick-up 3D modelling skills as the knowledge I have now is insufficient, In the past few days, I’ve started to learn Blender, which is a free software for 3D modelling and is great for my project, still, time is needed to hone the skills hence the long period of time allocated to learn these skills.

Initial Purchases (25th April – 27 April):
One of the biggest way to save money is to purchase them from china, which will take weeks for the item to arrive, hence it is really important to buy research about the parts required and buy them early to use the least money for the best result. Also, initial purchases is set to 25th because I will be presenting on that day, if no major changes were made, I could only really consider what to buy after that.

R1 Prototype (Software and Hardware) (2nd May- 8th July):
This will go hand in hand with skills acquiring, as I need to have a goal of what exactly to learn, it is the best to do while learning and learn while doing. R1 is the first Robot that I will be building of the set of 3, which will be the bare-bone of the Robot basic functionality and act as a confirmation to the general systems and parts requires to build R2 and R3.

R1 Movement System Finalization (18th June – 29 June):
As movement is really difficult task to achieve while concerning about the safety of the people and booth, (It is really easy to make something move, but it is much harder to make it move while not destroying things.)
so I gave more time for me to think about how I will achieve this.

Research 2 ( 20th June – 4th July):
When I think of robots, I will think of Japan, maybe its just me since I was influenced by the robotic culture of Japan when I was young, so I will travel to Japan during this period to experience their advancement of robotics first hand. (Place which I will visit :National Museum of Emerging Science and Innovation (Miraikan), Unicorn Gundam in Odaiba, bot at Haneda Airport, Henn-na Hotel, Robot Restaurant (not sure about this).

R1 Movement Prototype (4th July-23rd July):
Start to prototype right after I am back from Japan from experiencing their robots and hopefully get to see how they works in Japan.

R1+R2+R3 Concept Generation and Refinement( 4th July -27th July):
Since by this point I’ve already understand what parts R1 requires and already have the measurement of parts and sizes like motors and screen size, I could think about exactly how each Robots will look like as they will look different and have different functionality.

R1+R2+R3 Secondary Purchase (27th July – 29th July):
Knowing what parts each robot needs, I could finally purchase the basic parts for R2 and R3, plus the add-on function for all 3 robots(each robots have different functionality so require different parts)

3D modeling(Aesthetics) (27th July – 20th August):
This will be the final appearance for all 3 robots, 3D modelling done in blender.

R1 Prototype(Aesthetics+ Software + Hardware) (20th August – 24th Sep):
3D printing of all R1 modeled parts, fix them together and make sure the software and hardware works, if it doesn’t, edit and reprint of the parts.

R1 Prototype Trial and Testing ( 24th Sep – 1st October):
when all parts work together, test the robots and system in a location to make sure everything work as expected and fine tune.

R2 and R3 Prototype V1 (Software and Hardware) ( 1st October – 5th Nov):
Since the primary component and system of R2 and R3 is the same as the already working R1, these 2 robots will require lesser time and the main portion of this 2 robot will be in 3D printing and executing the different function in them.

User Interaction Trail and Testing ( 5th Nov – 19th Nov):
Testing and making sure that there are no major bugs in the system, touchscreen and functionality works well.

FYP Student’s Work Collection 1 (20th Nov – 1st Dec):
At this time, all 3 Robot can roughly work and I’ve already document these robots, so instead of just verbally telling them I will help them in making their FYP better, It will be more stimulating if I show them a system which already work and ask them to prepare a document for this system for their own benefit. (it will not be easy to ask people to do extra work, so I need to make sure that I sell my Idea to them really early*That’s 1 semester before the end of FYP* by making these robots cool and they will be losing out if their work is not in the systems.) Also, at this point they don’t have to send me any work and it is already the semester break so they have some time to think about what they want to prepare for the systems.

Booth System Conceptualization.( 20th Nov – 3rd Dec):
By this time, I should have the system of the robots working and I need to incorporate that into booth for our FYP, and this will probably be the time which we will know where our FYP will take place(in school or in public) and this will change how the booth system drastically, so it will be better to place this at the end of semester break.

Booth System Prototype (Software + Hardware) (3rd Dec – 14th Jan):
after the conceptualization, prototype will come next and I hope to have this done before the start of semester so that I will have the fully working prototype done and having the whole semester to polish my work, troubleshoot and bug fix.

Software and hardware Refinement (14th Jan – 1st April):
Software and Hardware refinement will take up most of the time as the real problem will usually emerge at this point of time where some shortcoming of the project will be apparent, also, there might be good suggestion/advice by people along the way and this will be the time to incorporate these wonderful suggestion into the project.

User Interaction Testing 2(1st April – 8th April):
Testing of the final system, to make sure all parts and component work as it should. if problem found, atleast there are time to replace these components.

FYP Student’s Work Collection 2( 1st April – 1st May):
The final collection of (Hopefully) all of the student’s work and adding them to the system once it’s collected. At the very least, there will be the basic information of every student which is uploaded to the FYP website.

Aesthetic Refinement( 8th April – 29th April):
The polishing and painting of the 3 robots and making of props/items for the booth. (when all software and hardware is working)

Booth Preparation & Stylization (1st May – 8th May):
Production of prints for booth, name cards/postcards and such.

FYP Show Preparation. (8th May – 10th May):
The actual preparation of the Booth and to bring the robots down to the exhibition area and set everything them up.

FYP SHOW( 10th May- 20th May):
Make sure the show runs smoothly, on-site repair if needed.