00:00:00
hello peers my name is Johnny I chose
00:00:03
the humanoid robot figure as my study
00:00:06
topic it's a very new company founded by
00:00:08
open AI called Figure robotics so the
00:00:11
series of this robot is called Figure so
00:00:14
don't get confused
00:00:16
there and moving along to the table of
00:00:19
content now this is followed by
00:00:22
professor's
00:00:24
requirements and then we move to
00:00:28
preface so uh like I stated here I think
00:00:32
the purpose of robots is to improve
00:00:34
human life not the other way around so I
00:00:37
think many years ago a British
00:00:41
Economist you know predicted kind of
00:00:43
predicted like 21st centu human would
00:00:45
only have to work three days in a week
00:00:49
and everybody will have plenty of food
00:00:51
on their table and people have a warm
00:00:53
house to
00:00:54
stay but obviously that didn't happen
00:00:57
the tools of Productions in humans hand
00:00:59
are much more advanced now than before
00:01:02
but the workload didn't decrease didn't
00:01:05
really decrease much um this video will
00:01:09
give you more detail if you want to look
00:01:11
at
00:01:11
it what uh figure 02 looks like
00:01:19
[Music]
00:01:30
a
00:01:31
[Music]
00:02:02
[Music]
00:02:13
[Music]
00:02:27
[Music]
00:02:34
[Music]
00:02:59
[Music]
00:03:19
all
00:03:22
[Music]
00:03:26
right next
00:03:28
slide um the history history of figure
00:03:31
is pretty new it was founded in 20
00:03:36
2022
00:03:38
and but what happened in the past two
00:03:41
years so in 2022 company introduced its
00:03:45
prototyp figure 01 a bip petto robot
00:03:50
only on January 2024 figure announced a
00:03:54
partnership with BMW in you know South
00:03:58
Carolina on August 2024 figure AI unveil
00:04:03
the next Generations of its humanoid
00:04:05
robot name figure
00:04:10
02 in terms of applications I figure
00:04:13
aims to work side by side with humans in
00:04:15
various industry enhancing productivity
00:04:18
and safety and you know figure amazingly
00:04:22
also offers speech to speech reasoning
00:04:24
and good
00:04:25
mobilities um in actuality the
00:04:28
applications of human robots or anness
00:04:31
you know there's so much possibility no
00:04:34
since figure can communicate and work in
00:04:36
the you know in the future he probably
00:04:38
be a housekeeper who also provides
00:04:40
emotional support a factory worker or a
00:04:44
uh prospector of dangerous area now I
00:04:48
recently heard that some of the humanoid
00:04:50
Bots can cook and they in their database
00:04:54
many recipes are are originated from
00:04:56
meling starts shfs so they they
00:05:00
probably they're capable to become very
00:05:03
good Chef or some of them will just
00:05:07
watch and learn you know you can learn
00:05:09
how your mom Cook and Cook you know
00:05:11
mom's taste you know mom's
00:05:13
meal
00:05:18
and you know also I always tell people
00:05:22
this compares to robot human bodies are
00:05:25
too fragile and
00:05:27
irreplaceable you know during the time
00:05:29
that was working for UPS the most common
00:05:32
complaints I heard from everybody is
00:05:34
pretty much everybody you know it was
00:05:36
people complaining they have back pain
00:05:38
back Aid and they their needs had
00:05:43
surgeries and I I don't think humans are
00:05:46
built for you know fastpac running
00:05:50
packages around or you know similar jobs
00:05:54
like construction or
00:05:57
something um a product like figures you
00:05:59
know can you know alleviate some of the
00:06:01
heavy lifting right so lower the
00:06:04
physical
00:06:06
demand uh in terms of
00:06:10
mechanics it figure humanoid robots
00:06:13
integrated Advanced AI model and you
00:06:16
know some what what do you call it like
00:06:20
flameless electric
00:06:22
BS so that it greatly enhance its
00:06:27
capabilities I think it's even better
00:06:29
than and hydraulic right now the
00:06:31
electric
00:06:37
motors um we're going talk about
00:06:42
structures you know like I said the
00:06:44
figure is built with electric motors
00:06:46
instead of hydraulic so the pro of that
00:06:50
it's very obvious it's more reliable and
00:06:53
requires less maintenance you know
00:06:55
speaking of AI
00:06:57
Powers you know I think
00:07:00
um open AI is the leader in the AI
00:07:03
industry and Nvidia is the leader of
00:07:05
computation power or say manufacturers
00:07:09
of advanced chip sets you know I believe
00:07:12
flamess torque Motors that I just talk
00:07:15
about are used to build high torque
00:07:18
flexible joints these are the
00:07:20
technologies that we didn't have just 10
00:07:23
years ago now these motors Motors are
00:07:26
designed to be energy efficient and the
00:07:30
absence of a frame allows for direction
00:07:32
strive now resulting in smooth and more
00:07:35
precise motion control now I think some
00:07:38
of these motors are capable of 27
00:07:42
directions of movement uh but I'm not
00:07:45
sure the the capability of the latest
00:07:48
model so I don't have much to say about
00:07:52
their AI model actually you know in
00:07:55
computational
00:07:57
power a lot of their information is open
00:07:59
Source but my knowledge it's just
00:08:01
insufficient so I don't want to mislead
00:08:04
anyone
00:08:06
here um how to train and build figures
00:08:09
series here it's clearly stated like
00:08:12
figure robotics has revolutionized the
00:08:14
field of humanoid robots robotics with
00:08:17
its Advanced figure series now these
00:08:20
robots incorporate frameless torque as I
00:08:23
mentioned it's it's flexible we didn't
00:08:25
have this before and with human scale
00:08:28
hand six RGB cameras and the you know
00:08:32
the excels in performance on high
00:08:34
position task fly electric F electric
00:08:37
you know power it run on batteries for
00:08:40
five hours long and it can you know
00:08:43
carry 20 kg or 44
00:08:46
pounds walking speed it's pretty slow
00:08:49
but uh in the future it's likely to
00:08:52
increase a lot and like I said the AI
00:08:55
integration is the most important part
00:08:58
for this figure series of
00:09:03
robots it's you know first it train with
00:09:07
partnership with AI leaders like open Ai
00:09:09
and Nvidia and all AI models train on
00:09:12
Nvidia h100
00:09:15
gpus it's it's very expensive GPU
00:09:19
probably over you know 10K for for a
00:09:22
piece probably I'm not sure but
00:09:26
um yeah it's the Leading Edge
00:09:32
so
00:09:33
evaluation so I said this when the
00:09:35
quarter started you know the
00:09:42
pros Provide support to Elder know like
00:09:45
I said I mentioned my grand uh
00:09:47
grandfather before he passed away it it
00:09:49
was painful for
00:09:51
everybody and human don't
00:09:55
judge so it's it's better for the
00:09:57
patient it's better for the family
00:09:58
members
00:10:00
better for the medical
00:10:02
staffs and now also hoping to see I hope
00:10:07
if if this can you know um human robots
00:10:10
start entering the societies I wish to
00:10:13
see in the medical field first because
00:10:16
the cost of hospitalization in the US is
00:10:19
just too too expensive right now now
00:10:23
people moving to the second Point High
00:10:26
position and without fatigue so this is
00:10:28
very important important like people
00:10:31
probably can focus on their job and work
00:10:33
hard for eight hours now after that
00:10:37
especially in physical
00:10:39
work after 8 hours people likely to make
00:10:42
more mistakes and no fear and
00:10:49
anger figure Bots can go into some
00:10:52
dangerous Stone like cave and
00:10:55
underground tunnels which human would
00:10:58
probably not attempt
00:10:59
it's just not worth a human life it
00:11:04
takes people 18 years to grow up but
00:11:07
robot your manufacturing on the
00:11:08
production lines just different so
00:11:11
reduce workplace injury and fatality
00:11:13
it's also important like I said earlier
00:11:17
uh I don't think human are made for
00:11:19
heavy lifting it's it's just not like I
00:11:23
couple years ago I hurt my back on a job
00:11:27
when I was working at UPS and
00:11:30
it the problem still bothers me if I
00:11:33
sleep on my bed for over 10 hours my bat
00:11:36
get got you know my back just start
00:11:39
feeling stiff it's just not comfortable
00:11:41
if I try to continue to
00:11:43
sleep and my match is actually pretty
00:11:46
good before when I had my old mattress I
00:11:48
had to you know just threw it away
00:11:50
otherwise I I couldn't even sleep longer
00:11:52
than six hours cons you know let's let's
00:11:55
talk about cons yeah some people would
00:11:58
lose their job
00:12:00
but they should be trained to do a
00:12:02
better jobs
00:12:04
right and the company can do that the
00:12:06
company can train their their employees
00:12:08
to do a better job or the government
00:12:10
should perform a task and once we step
00:12:13
into the big digiti database we do lose
00:12:18
our
00:12:19
privacy and we really can't hide
00:12:22
anything now we just need to better
00:12:24
regulate AI companies and data managing
00:12:27
firms and the elment of the figure is
00:12:31
very expensive this is this is common
00:12:34
you know when we got to hire a lot
00:12:37
of um senior
00:12:40
Engineers AI programmers and all that
00:12:42
it's always going to be
00:12:44
expensive and that's how that's how it
00:12:47
is
00:12:50
but and then over Reliance on robots may
00:12:55
cause people to lose hands on skill um
00:12:58
this is common scene one example is when
00:13:01
people use GPS to navigate all the time
00:13:03
then they learn the route much slower as
00:13:07
well as people don't ever cook they
00:13:10
would never learn how to
00:13:11
cook and these are the pros and cons I
00:13:14
can come up
00:13:15
with um for future outlook the market
00:13:19
growth market growth for humanoid robots
00:13:22
are significant astonishing from 1.6
00:13:26
billion in 2022 to 24 14.4 billions by
00:13:31
2032 so you know getting a job or uh
00:13:35
studying for this topic always
00:13:38
gooded yeah
00:13:40
Integrations
00:13:43
um for this part for AI model I can't
00:13:47
really say much you know I learned
00:13:49
Python and Java but I I don't really
00:13:51
have any work experience I don't know
00:13:54
how the real world works but what I
00:13:57
learned online is AI model will be fine
00:13:59
tuned custom
00:14:02
multimodels models multi multimodel
00:14:05
models on humano robot data so probably
00:14:10
that just mean integrated models like
00:14:13
big models integrated together to make
00:14:16
the humanite Bots uh cope or to fit more
00:14:21
area more work
00:14:24
places and uh thank you for listening
00:14:27
here's my conclusions figure series or
00:14:30
humanoid robots are based on a data
00:14:32
engine that powers learning for a fleet
00:14:35
at affordable humanoid I believe that
00:14:38
the advancement of human technology will
00:14:40
ultimately benefit the majority what we
00:14:44
should do is to stay optimistic and keep
00:14:46
our minds open to learn new things every
00:14:51
day and here are the sources that I use
00:14:55
for this
00:14:56
PowerPoint thank you so much peers and
00:14:59
thank you professor
