Robots – From Programming to Learning | Torsten Kröger | TEDxKIT

so as as you all know robotics

so as as you all know robotics artificial and tell

artificial and tell

artificial and tell learning this all buzzwords

learning this all buzzwords

learning this all buzzwords can not only find in research and

can not only find in research and

can not only find in research and science anymore you can find them more

science anymore you can find them more

science anymore you can find them more and more in public media so when there

and more in public media so when there

and more in public media so when there is a phase where you have this this big

is a phase where you have this this big

is a phase where you have this this big hype about a domain it is always good to

hype about a domain it is always good to

hype about a domain it is always good to start with the definition so what is

start with the definition so what is

start with the definition so what is artificial intelligence machine learning

artificial intelligence machine learning

artificial intelligence machine learning means software software algorithms that

means software software algorithms that

means software software algorithms that learn from data robotics means we have a

learn from data robotics means we have a

learn from data robotics means we have a connection between a computer system a

connection between a computer system a

connection between a computer system a digital system that can physically

digital system that can physically

digital system that can physically interact with the with the environment

interact with the with the environment

interact with the with the environment and we’re going to talk about that much

and we’re going to talk about that much

and we’re going to talk about that much much more in detail today but taking a

much more in detail today but taking a

much more in detail today but taking a step back so there are self-driving cars

step back so there are self-driving cars

step back so there are self-driving cars I’m sure that every one of us has

I’m sure that every one of us has

I’m sure that every one of us has thought about that looking at Apple Siri

thought about that looking at Apple Siri

thought about that looking at Apple Siri Google’s assistant Amazon’s Alexa so

Google’s assistant Amazon’s Alexa so

Google’s assistant Amazon’s Alexa so we’re seeing more and more of these

we’re seeing more and more of these

we’re seeing more and more of these technologies entering our everyday lives

technologies entering our everyday lives

technologies entering our everyday lives and what makes this different so when it

and what makes this different so when it

and what makes this different so when it comes artificial intelligence what

comes artificial intelligence what

comes artificial intelligence what computer software used to be imagine you

computer software used to be imagine you

computer software used to be imagine you had three colors white grey and black

had three colors white grey and black

had three colors white grey and black and you would ask a computer program

and you would ask a computer program

and you would ask a computer program what color is this white grey or black

what color is this white grey or black

what color is this white grey or black there’s a unique answer to this if you

there’s a unique answer to this if you

there’s a unique answer to this if you show a photograph to an algorithm and

show a photograph to an algorithm and

show a photograph to an algorithm and you would ask who is on the photograph

you would ask who is on the photograph

you would ask who is on the photograph then the answer might be this could be

then the answer might be this could be

then the answer might be this could be Linda but maybe it’s some someone else

Linda but maybe it’s some someone else

Linda but maybe it’s some someone else or this could be Chris but it could be

or this could be Chris but it could be

or this could be Chris but it could be someone else but the same is true for us

someone else but the same is true for us

someone else but the same is true for us humans right if I would show you a

humans right if I would show you a

humans right if I would show you a photograph in many cases assuming you

photograph in many cases assuming you

photograph in many cases assuming you know that person you can tell who is on

know that person you can tell who is on

know that person you can tell who is on the photograph but in many cases also

the photograph but in many cases also

the photograph but in many cases also you you’re not really sure it could be

you you’re not really sure it could be

you you’re not really sure it could be that person but could be someone else so

that person but could be someone else so

that person but could be someone else so we’re seeing more and more of these

we’re seeing more and more of these

we’re seeing more and more of these breakthroughs in semantic understanding

breakthroughs in semantic understanding

breakthroughs in semantic understanding speech recognition speech translation

speech recognition speech translation

speech recognition speech translation now looking at robotics let’s take a

now looking at robotics let’s take a

now looking at robotics let’s take a step back to 1997 when IBM’s deep blue

step back to 1997 when IBM’s deep blue

step back to 1997 when IBM’s deep blue bet Garry Kasparov in chess this is what

bet Garry Kasparov in chess this is what

bet Garry Kasparov in chess this is what we we call it artificial intelligence at

we we call it artificial intelligence at

we we call it artificial intelligence at that point and I’m sure that you’re for

that point and I’m sure that you’re for

that point and I’m sure that you’re for of alphago are much more recently

of alphago are much more recently

of alphago are much more recently however I have not yet seen a computer

however I have not yet seen a computer

however I have not yet seen a computer system or robot that can take a

system or robot that can take a

system or robot that can take a box of chess figures take each

box of chess figures take each

box of chess figures take each individual figure put it on the board

individual figure put it on the board

individual figure put it on the board and physically play chess with me so

and physically play chess with me so

and physically play chess with me so let’s let’s talk about robotics what is

let’s let’s talk about robotics what is

let’s let’s talk about robotics what is what is robotics robotics is the

what is robotics robotics is the

what is robotics robotics is the intelligent connection between

intelligent connection between

intelligent connection between perception and action so what that means

perception and action so what that means

perception and action so what that means is on the right side you can see a robot

is on the right side you can see a robot

is on the right side you can see a robot and the robot has is installed in some

and the robot has is installed in some

and the robot has is installed in some environment works in some environment

environment works in some environment

environment works in some environment but it’s also about physical contact

but it’s also about physical contact

but it’s also about physical contact having the system making physical

having the system making physical

having the system making physical physical contact with the environment

physical contact with the environment

physical contact with the environment possibly with the human and the robot

possibly with the human and the robot

possibly with the human and the robot consists of actuators in many cases

consists of actuators in many cases

consists of actuators in many cases electrical motors but there are many

electrical motors but there are many

electrical motors but there are many other actuators and then it consists of

other actuators and then it consists of

other actuators and then it consists of sensors so it can sense its own position

sensors so it can sense its own position

sensors so it can sense its own position it might have a camera for visual

it might have a camera for visual

it might have a camera for visual perception it might have a forced walk

perception it might have a forced walk

perception it might have a forced walk sensor or skin for tactile perception

sensor or skin for tactile perception

sensor or skin for tactile perception and then we have a a control system and

and then we have a a control system and

and then we have a a control system and what I want to focus on today is this

what I want to focus on today is this

what I want to focus on today is this control system so let’s let me let me

control system so let’s let me let me

control system so let’s let me let me start with something that I’ve done

start with something that I’ve done

start with something that I’ve done quite a while ago and this is one of the

quite a while ago and this is one of the

quite a while ago and this is one of the first projects that I did and in

first projects that I did and in

first projects that I did and in robotics what you can see here is a

robotics what you can see here is a

robotics what you can see here is a robot inserting a battery into a cell

robot inserting a battery into a cell

robot inserting a battery into a cell phone the robot is equipped with a four

phone the robot is equipped with a four

phone the robot is equipped with a four stroke sensor so it can can feel but it

stroke sensor so it can can feel but it

stroke sensor so it can can feel but it cannot see the phone and it doesn’t

cannot see the phone and it doesn’t

cannot see the phone and it doesn’t exactly know where the phone is so you

exactly know where the phone is so you

exactly know where the phone is so you can see that it is it is displaced right

can see that it is it is displaced right

can see that it is it is displaced right so that you know we don’t know the

so that you know we don’t know the

so that you know we don’t know the position and orientation of the phone so

position and orientation of the phone so

position and orientation of the phone so we’re just feeling like blindly trying

we’re just feeling like blindly trying

we’re just feeling like blindly trying to insert the battery into the phone and

to insert the battery into the phone and

to insert the battery into the phone and what you can see on the right hand side

what you can see on the right hand side

what you can see on the right hand side is a graph so the graph is static and

is a graph so the graph is static and

is a graph so the graph is static and each node in this graph is a command for

each node in this graph is a command for

each node in this graph is a command for the robot so depending on the sensor

the robot so depending on the sensor

the robot so depending on the sensor signals from the robot we take a a path

signals from the robot we take a a path

signals from the robot we take a a path through this graph that really depends

through this graph that really depends

through this graph that really depends on what is happening in the robot work

on what is happening in the robot work

on what is happening in the robot work cell now when we switch from one node to

cell now when we switch from one node to

cell now when we switch from one node to another we don’t know in what state we

another we don’t know in what state we

another we don’t know in what state we are in so that means we need to be able

are in so that means we need to be able

are in so that means we need to be able to switch to react instantaneously in

to switch to react instantaneously in

to switch to react instantaneously in this case within like one millisecond we

this case within like one millisecond we

this case within like one millisecond we need to change from one behavior to

need to change from one behavior to

need to change from one behavior to another so

another so

another so this is comparable to the human reflex

this is comparable to the human reflex

this is comparable to the human reflex you can see me soldering and I’m burning

you can see me soldering and I’m burning

you can see me soldering and I’m burning my fingers and these these reflexes are

my fingers and these these reflexes are

my fingers and these these reflexes are very often triggered in our spinal cord

very often triggered in our spinal cord

very often triggered in our spinal cord or in our surah bellum and we react to

or in our surah bellum and we react to

or in our surah bellum and we react to pain or something unforeseen right away

pain or something unforeseen right away

pain or something unforeseen right away and very often actually not very often

and very often actually not very often

and very often actually not very often we we really first react and then

we we really first react and then

we we really first react and then afterwards we consciously notice oh I

afterwards we consciously notice oh I

afterwards we consciously notice oh I just brought my fingers so we use this

just brought my fingers so we use this

just brought my fingers so we use this technology and we implemented a Jenga

technology and we implemented a Jenga

technology and we implemented a Jenga playing robot so Jang has a parlor game

playing robot so Jang has a parlor game

playing robot so Jang has a parlor game so many of you know that it consists of

so many of you know that it consists of

so many of you know that it consists of a tower of wooden bricks and the goal is

a tower of wooden bricks and the goal is

a tower of wooden bricks and the goal is to pick a brick take it out and put it

to pick a brick take it out and put it

to pick a brick take it out and put it back on the top of the tower and the

back on the top of the tower and the

back on the top of the tower and the robot that we used to play this game was

robot that we used to play this game was

robot that we used to play this game was equipped with position sensors in the

equipped with position sensors in the

equipped with position sensors in the joints so the robot knew where it was

joints so the robot knew where it was

joints so the robot knew where it was it had a forced walk sensor so that it

it had a forced walk sensor so that it

it had a forced walk sensor so that it could feel it had a distance sensor and

could feel it had a distance sensor and

could feel it had a distance sensor and it had two cameras and what you know can

it had two cameras and what you know can

it had two cameras and what you know can can see we’re now about to to grasp this

can see we’re now about to to grasp this

can see we’re now about to to grasp this one brick of wood and while we’re now

one brick of wood and while we’re now

one brick of wood and while we’re now trying to to pull out this this break

trying to to pull out this this break

trying to to pull out this this break we’re eliminating all transfers of

we’re eliminating all transfers of

we’re eliminating all transfers of forces and torques and we do that to

forces and torques and we do that to

forces and torques and we do that to prevent the tower from toppling and we

prevent the tower from toppling and we

prevent the tower from toppling and we really have to do this in a very very

really have to do this in a very very

really have to do this in a very very careful careful manner and this again is

careful careful manner and this again is

careful careful manner and this again is implemented using the same primitives

implemented using the same primitives

implemented using the same primitives there’s the same graph principle that

there’s the same graph principle that

there’s the same graph principle that you that you’ve seen before however now

you that you’ve seen before however now

you that you’ve seen before however now this graph is much more complex um you

this graph is much more complex um you

this graph is much more complex um you can see a view from from the cameras and

can see a view from from the cameras and

can see a view from from the cameras and we were actually cheating a little bit

we were actually cheating a little bit

we were actually cheating a little bit so we we colored all of the blocks with

so we we colored all of the blocks with

so we we colored all of the blocks with black and white color so that it’s not a

black and white color so that it’s not a

black and white color so that it’s not a computer vision problem so we simplified

computer vision problem so we simplified

computer vision problem so we simplified our task that way you can see the the

our task that way you can see the the

our task that way you can see the the force torque sensor to really get the

force torque sensor to really get the

force torque sensor to really get the the tactile perception you can see it in

the tactile perception you can see it in

the tactile perception you can see it in on the robot and the distance sensor as

on the robot and the distance sensor as

on the robot and the distance sensor as well now taking a step forward

well now taking a step forward

well now taking a step forward programming here was was really a

programming here was was really a

programming here was was really a challenge

challenge

challenge and looking at the control system what

and looking at the control system what

and looking at the control system what could we do to kind of make the control

could we do to kind of make the control

could we do to kind of make the control system more intelligent the the green

system more intelligent the the green

system more intelligent the the green block here on the left so we could

block here on the left so we could

block here on the left so we could imagine to put a aun in there right so

imagine to put a aun in there right so

imagine to put a aun in there right so the human is Telly operating the robot

the human is Telly operating the robot

the human is Telly operating the robot and if the human also gets a sense of

and if the human also gets a sense of

and if the human also gets a sense of touch then this is what we call haptics

touch then this is what we call haptics

touch then this is what we call haptics so the human can haptically telle

so the human can haptically telle

so the human can haptically telle operate the robot so the human can feel

operate the robot so the human can feel

operate the robot so the human can feel what the robot is feeling and here’s one

what the robot is feeling and here’s one

what the robot is feeling and here’s one one example this is something that we

one example this is something that we

one example this is something that we did at Stanford University you can see

did at Stanford University you can see

did at Stanford University you can see this wealth on the right and the robot

this wealth on the right and the robot

this wealth on the right and the robot as well and this is a haptic device a

as well and this is a haptic device a

as well and this is a haptic device a haptic device is a imagine a joystick

haptic device is a imagine a joystick

haptic device is a imagine a joystick and you have force feedback but with

and you have force feedback but with

and you have force feedback but with this joystick you you get force feedback

this joystick you you get force feedback

this joystick you you get force feedback and up to six degrees of freedom so you

and up to six degrees of freedom so you

and up to six degrees of freedom so you can feel exactly what the robot is

can feel exactly what the robot is

can feel exactly what the robot is feeling so in this case the human

feeling so in this case the human

feeling so in this case the human operator can actually see the valves but

operator can actually see the valves but

operator can actually see the valves but of course there could also be a camera

of course there could also be a camera

of course there could also be a camera so that the operator would only see the

so that the operator would only see the

so that the operator would only see the the camera image No

the camera image No

the camera image No obviously that doesn’t scale right we

obviously that doesn’t scale right we

obviously that doesn’t scale right we would need one human per robot so would

would need one human per robot so would

would need one human per robot so would be easy to program because it’s done by

be easy to program because it’s done by

be easy to program because it’s done by the human but we can scale that so we

the human but we can scale that so we

the human but we can scale that so we need something else a much more

need something else a much more

need something else a much more interesting aspect is actually putting

interesting aspect is actually putting

interesting aspect is actually putting the human not at the channel operation

the human not at the channel operation

the human not at the channel operation seat but have the human and the robot

seat but have the human and the robot

seat but have the human and the robot interact maybe physically interact so

interact maybe physically interact so

interact maybe physically interact so here’s another project where you can see

here’s another project where you can see

here’s another project where you can see the robot and the human physically

the robot and the human physically

the robot and the human physically interacting this actually was a student

interacting this actually was a student

interacting this actually was a student project at Stanford University I’m using

project at Stanford University I’m using

project at Stanford University I’m using this this framework that I mentioned

this this framework that I mentioned

this this framework that I mentioned earlier where we are able to generate

earlier where we are able to generate

earlier where we are able to generate notions within less than one millisecond

notions within less than one millisecond

notions within less than one millisecond comparing this is what what is

comparing this is what what is

comparing this is what what is comparable to the human reflex and here

comparable to the human reflex and here

comparable to the human reflex and here the students were able to implement this

the students were able to implement this

the students were able to implement this in in less than three weeks and it’s

in in less than three weeks and it’s

in in less than three weeks and it’s actually it looks complex but it’s it’s

actually it looks complex but it’s it’s

actually it looks complex but it’s it’s very very simple so the robot moves

very very simple so the robot moves

very very simple so the robot moves towards the human sort and the moment we

towards the human sort and the moment we

towards the human sort and the moment we detect the contact within one

detect the contact within one

detect the contact within one millisecond the robot starts recoiling

millisecond the robot starts recoiling

millisecond the robot starts recoiling exactly what a human would do as well

exactly what a human would do as well

exactly what a human would do as well and here’s the defense mode also very

and here’s the defense mode also very

and here’s the defense mode also very simple all the the only thing that the

simple all the the only thing that the

simple all the the only thing that the human and sorry that the robot tries to

human and sorry that the robot tries to

human and sorry that the robot tries to do is to put it sort perpendicular to

do is to put it sort perpendicular to

do is to put it sort perpendicular to the human sort you can see that again so

the human sort you can see that again so

the human sort you can see that again so you can see me fighting with the robot

you can see me fighting with the robot

you can see me fighting with the robot and all the robot is trying to do is to

and all the robot is trying to do is to

and all the robot is trying to do is to put it it’s sort perpendicular to mine

put it it’s sort perpendicular to mine

put it it’s sort perpendicular to mine and you can see the 3d camera there and

and you can see the 3d camera there and

and you can see the 3d camera there and also the camera image in in the

also the camera image in in the

also the camera image in in the background

background

background now looking at this this actually looks

now looking at this this actually looks

now looking at this this actually looks dangerous right and imagine you would

dangerous right and imagine you would

dangerous right and imagine you would have humans and and robots interacting

have humans and and robots interacting

have humans and and robots interacting sharing the same Rock space so safety is

sharing the same Rock space so safety is

sharing the same Rock space so safety is something physical safety for the human

something physical safety for the human

something physical safety for the human is something that really needs to have

is something that really needs to have

is something that really needs to have the highest priority whatsoever so we

the highest priority whatsoever so we

the highest priority whatsoever so we don’t want to harm to injure or even to

don’t want to harm to injure or even to

don’t want to harm to injure or even to kill a human right so this has to be um

kill a human right so this has to be um

kill a human right so this has to be um the the foremost priority for for

the the foremost priority for for

the the foremost priority for for roboticists when they work on

roboticists when they work on

roboticists when they work on applications where robots and humans

applications where robots and humans

applications where robots and humans share the same workspace so we then

share the same workspace so we then

share the same workspace so we then worked on on an algorithm that allows

worked on on an algorithm that allows

worked on on an algorithm that allows the robot to prevent from collisions so

the robot to prevent from collisions so

the robot to prevent from collisions so you can see on the right-hand side here

you can see on the right-hand side here

you can see on the right-hand side here a a depth image and based on the depth

a a depth image and based on the depth

a a depth image and based on the depth image we can detect the human we can

image we can detect the human we can

image we can detect the human we can detect obstacles and we then compute a

detect obstacles and we then compute a

detect obstacles and we then compute a velocity vector so we compute a point in

velocity vector so we compute a point in

velocity vector so we compute a point in space and the velocity vector is

space and the velocity vector is

space and the velocity vector is pointing into the opposite direction of

pointing into the opposite direction of

pointing into the opposite direction of the human and to the opposite direction

the human and to the opposite direction

the human and to the opposite direction of of the obstacle and this way we can

of of the obstacle and this way we can

of of the obstacle and this way we can compute a motion in real time that

compute a motion in real time that

compute a motion in real time that allows the robot to move away from the

allows the robot to move away from the

allows the robot to move away from the obstacle in the best possible way as

obstacle in the best possible way as

obstacle in the best possible way as fast as physically possible so the robot

fast as physically possible so the robot

fast as physically possible so the robot tries everything it’s possibly Candace

tries everything it’s possibly Candace

tries everything it’s possibly Candace physically can to prevent from that

physically can to prevent from that

physically can to prevent from that collision so that is one trend but even

collision so that is one trend but even

collision so that is one trend but even here programming again this is very very

here programming again this is very very

here programming again this is very very expensive

expensive

expensive there is another development in the

there is another development in the

there is another development in the domain of Robotics research and that’s

domain of Robotics research and that’s

domain of Robotics research and that’s based on on unmoral it’s a lot of

based on on unmoral it’s a lot of

based on on unmoral it’s a lot of engineering we put a lot of

engineering we put a lot of

engineering we put a lot of and thought into the controllers into

and thought into the controllers into

and thought into the controllers into the motion planners and when I worked at

the motion planners and when I worked at

the motion planners and when I worked at Google I had the honor and the pleasure

Google I had the honor and the pleasure

Google I had the honor and the pleasure to also collaborate with the team at

to also collaborate with the team at

to also collaborate with the team at Boston Dynamics and I would like to give

Boston Dynamics and I would like to give

Boston Dynamics and I would like to give great credits to to them for the for the

great credits to to them for the for the

great credits to to them for the for the videos we’re seeing here this robot is

videos we’re seeing here this robot is

videos we’re seeing here this robot is called Atlas it’s a humanoid robot and

called Atlas it’s a humanoid robot and

called Atlas it’s a humanoid robot and this robot is fully force and torque

this robot is fully force and torque

this robot is fully force and torque controlled and what they were able to do

controlled and what they were able to do

controlled and what they were able to do here is to have the robot being able to

here is to have the robot being able to

here is to have the robot being able to walk alongside with humans on very

walk alongside with humans on very

walk alongside with humans on very uneven terrain – to manipulate objects

uneven terrain – to manipulate objects

uneven terrain – to manipulate objects like here and there’s another robot

like here and there’s another robot

like here and there’s another robot called called spot so this spot is the

called called spot so this spot is the

called called spot so this spot is the UM

UM

UM splat mini to be precise it’s a

splat mini to be precise it’s a

splat mini to be precise it’s a quadrotor robot within with an arm on

quadrotor robot within with an arm on

quadrotor robot within with an arm on top of it and what you can see here in

top of it and what you can see here in

top of it and what you can see here in this demonstration is actually whole

this demonstration is actually whole

this demonstration is actually whole body controlled so we keep the position

body controlled so we keep the position

body controlled so we keep the position of the end effector where it is and then

of the end effector where it is and then

of the end effector where it is and then we move move around the latest version

we move move around the latest version

we move move around the latest version of Atlas can even jump and this is

of Atlas can even jump and this is

of Atlas can even jump and this is really an amazing achievement they’ve

really an amazing achievement they’ve

really an amazing achievement they’ve just was implemented very recently and

just was implemented very recently and

just was implemented very recently and it’s not only about jumping and looking

it’s not only about jumping and looking

it’s not only about jumping and looking at some of you might have seen this

at some of you might have seen this

at some of you might have seen this already

so this is really an amazing engineering

so this is really an amazing engineering achievement and and credit to the team

achievement and and credit to the team

achievement and and credit to the team at Boston Dynamics at this point however

at Boston Dynamics at this point however

at Boston Dynamics at this point however in all three cases the graph that you’ve

in all three cases the graph that you’ve

in all three cases the graph that you’ve seen human robot collaboration and the

seen human robot collaboration and the

seen human robot collaboration and the the latest videos it requires a lot of

the latest videos it requires a lot of

the latest videos it requires a lot of skills expertise and experience to

skills expertise and experience to

skills expertise and experience to program these robots so how can we

program these robots so how can we

program these robots so how can we overcome this this problem you’re right

overcome this this problem you’re right

overcome this this problem you’re right we cannot have like people who are

we cannot have like people who are

we cannot have like people who are highly educated in robotics to program

highly educated in robotics to program

highly educated in robotics to program all of these robots it will not work it

all of these robots it will not work it

all of these robots it will not work it will not scale so this is where machine

will not scale so this is where machine

will not scale so this is where machine learning might come to help and this is

learning might come to help and this is

learning might come to help and this is ongoing research in many research labs

ongoing research in many research labs

ongoing research in many research labs in in the world including the ones here

in in the world including the ones here

in in the world including the ones here in in Karlsruhe and machine learning as

in in Karlsruhe and machine learning as

in in Karlsruhe and machine learning as I mentioned means software algorithms

I mentioned means software algorithms

I mentioned means software algorithms that learn from data now data is

that learn from data now data is

that learn from data now data is something that is sometimes hard to find

something that is sometimes hard to find

something that is sometimes hard to find in robotics so we can find a lot of text

in robotics so we can find a lot of text

in robotics so we can find a lot of text and images in the internet to train

and images in the internet to train

and images in the internet to train algorithms this is why they have become

algorithms this is why they have become

algorithms this is why they have become very very trained in speech recognition

very very trained in speech recognition

very very trained in speech recognition image recognition however they’re not

image recognition however they’re not

image recognition however they’re not too many websites that show motor

too many websites that show motor

too many websites that show motor currents motor torques joint torques

currents motor torques joint torques

currents motor torques joint torques joint positions right we just don’t have

joint positions right we just don’t have

joint positions right we just don’t have enough enough of their data so the

enough enough of their data so the

enough enough of their data so the problem we have here is we actually need

problem we have here is we actually need

problem we have here is we actually need a lot of data it’s really about data and

a lot of data it’s really about data and

a lot of data it’s really about data and and we come to data driven robotics and

and we come to data driven robotics and

and we come to data driven robotics and one experiment that we did it at Google

one experiment that we did it at Google

one experiment that we did it at Google awhile ago this was last year we had 14

awhile ago this was last year we had 14

awhile ago this was last year we had 14 robots trying to grasp objects and we

robots trying to grasp objects and we

robots trying to grasp objects and we did there for two months there’s a

did there for two months there’s a

did there for two months there’s a method that is called reinforcement

method that is called reinforcement

method that is called reinforcement learning and that means the robot is is

learning and that means the robot is is

learning and that means the robot is is taking an action and if it does a a good

taking an action and if it does a a good

taking an action and if it does a a good action it gets a reward for it so in

action it gets a reward for it so in

action it gets a reward for it so in this case the reward was the success of

this case the reward was the success of

this case the reward was the success of grasping an object and we did that as I

grasping an object and we did that as I

grasping an object and we did that as I mentioned for about two months that was

mentioned for about two months that was

mentioned for about two months that was a little more than 800,000 grasps and

a little more than 800,000 grasps and

a little more than 800,000 grasps and there was one thing that happened it was

there was one thing that happened it was

there was one thing that happened it was truly amazing and surprising to us so

truly amazing and surprising to us so

truly amazing and surprising to us so look at this scene you can see the

look at this scene you can see the

look at this scene you can see the stapler

stapler

stapler the yellow DUPLO break here and now in

the yellow DUPLO break here and now in

the yellow DUPLO break here and now in the next scene here the robot is trying

the next scene here the robot is trying

the next scene here the robot is trying to grasp the yellow brick and what it

to grasp the yellow brick and what it

to grasp the yellow brick and what it does is it pushes away the stapler to

does is it pushes away the stapler to

does is it pushes away the stapler to then grasp the object and there was no

then grasp the object and there was no

then grasp the object and there was no engineer that wrote a single line of

engineer that wrote a single line of

engineer that wrote a single line of code to do this this was truly learned

code to do this this was truly learned

code to do this this was truly learned by the system and when we’ve seen that

by the system and when we’ve seen that

by the system and when we’ve seen that this was a big surprise and a big

this was a big surprise and a big

this was a big surprise and a big success to thee to the entire team there

success to thee to the entire team there

success to thee to the entire team there now this is truly amazing however even

now this is truly amazing however even

now this is truly amazing however even that doesn’t scale because this works

that doesn’t scale because this works

that doesn’t scale because this works for this robot with this camera with

for this robot with this camera with

for this robot with this camera with this grouper if you change the setup the

this grouper if you change the setup the

this grouper if you change the setup the neural network that was trained will not

neural network that was trained will not

neural network that was trained will not work anymore so machine learning is no

work anymore so machine learning is no

work anymore so machine learning is no magic and we have to be aware of it it’s

magic and we have to be aware of it it’s

magic and we have to be aware of it it’s a tool and it just means it’s a set of

a tool and it just means it’s a set of

a tool and it just means it’s a set of algorithms that learn from data now

algorithms that learn from data now

algorithms that learn from data now taking us stop here if if you are

taking us stop here if if you are

taking us stop here if if you are interested in in this domain and in

interested in in this domain and in

interested in in this domain and in robotics there’s a fantastic a book and

robotics there’s a fantastic a book and

robotics there’s a fantastic a book and also a website where you can find

also a website where you can find

also a website where you can find hundreds of such videos and technologies

hundreds of such videos and technologies

hundreds of such videos and technologies with descriptions it’s all openly

with descriptions it’s all openly

with descriptions it’s all openly available and if you take a look at this

available and if you take a look at this

available and if you take a look at this this book and website I’m sure you will

this book and website I’m sure you will

this book and website I’m sure you will find more interesting information about

find more interesting information about

find more interesting information about these technologies and now concluding

these technologies and now concluding

these technologies and now concluding looking a little bit into the future

looking a little bit into the future

looking a little bit into the future of what’s what’s gonna happen next so

of what’s what’s gonna happen next so

of what’s what’s gonna happen next so we’re seeing almost weekly right now

we’re seeing almost weekly right now

we’re seeing almost weekly right now technical breakthroughs in visual

technical breakthroughs in visual

technical breakthroughs in visual perception and and speech recognition

perception and and speech recognition

perception and and speech recognition however the challenges that we are still

however the challenges that we are still

however the challenges that we are still facing and where we do not yet have good

facing and where we do not yet have good

facing and where we do not yet have good solutions is safe actuation having

solutions is safe actuation having

solutions is safe actuation having robots and humans working together in a

robots and humans working together in a

robots and humans working together in a safe way having robots physically

safe way having robots physically

safe way having robots physically manipulating the environment and having

manipulating the environment and having

manipulating the environment and having robots and humans working together

robots and humans working together

robots and humans working together sharing the same workspace and doing

sharing the same workspace and doing

sharing the same workspace and doing that in a in a safe way so to conclude

that in a in a safe way so to conclude

that in a in a safe way so to conclude this is what we’ve seen before so the

this is what we’ve seen before so the

this is what we’ve seen before so the intelligent connection between

intelligent connection between

intelligent connection between perception and action safety is one very

perception and action safety is one very

perception and action safety is one very important very fundamental aspect and

important very fundamental aspect and

important very fundamental aspect and data so this is where we are going to

data so this is where we are going to

data so this is where we are going to see in the next decade more and more

see in the next decade more and more

see in the next decade more and more breakthroughs as well and I hope that I

breakthroughs as well and I hope that I

breakthroughs as well and I hope that I will be able to contribute that further

will be able to contribute that further

will be able to contribute that further but I would also like to say that we

but I would also like to say that we

but I would also like to say that we should not be afraid of this future

should not be afraid of this future

should not be afraid of this future robots are not gonna become self-aware

robots are not gonna become self-aware

robots are not gonna become self-aware they will help us support us assist us

they will help us support us assist us

they will help us support us assist us at work and in everyday life thank you

at work and in everyday life thank you

at work and in everyday life thank you so much

so much

so much [Applause]