Why Don’t We Have Better Robots Yet? | Ken Goldberg | TED
Summary
TLDRIn this talk, the speaker humorously addresses the gap between our expectations of home robots and the reality, highlighting the challenges of robot manipulation. Despite advancements in technology like tricorders and satellites, home robots capable of complex tasks remain elusive. The speaker, a researcher at UC Berkeley, delves into Moravec's paradox, explaining why simple human tasks are difficult for robots. They discuss hardware and software hurdles, including the unreliability of robot hands and the complexities of perception and physics. The speaker shares their work on AI and deep learning to improve robot grasping in e-commerce, showcasing the progress with Dex-net and Ambi Robotics. They also touch on ongoing research for home robots, including handling deformable objects and everyday tasks like folding laundry, emphasizing the ongoing efforts despite the slow pace of advancement.
Takeaways
- 🤖 The speaker humorously points out the gap between the advanced technology we have (like tricorders and satellites) and the lack of helpful home robots.
- 🔍 The reality of robotics is often far less impressive than what's seen in popular culture, with many robots failing simple tasks.
- 🤔 Moravec's paradox is introduced as a key concept, highlighting that tasks easy for robots are hard for humans, and vice versa.
- 🤲 The development of robot hands is discussed, with a preference for simple, reliable, and cost-effective designs like the parallel jaw gripper.
- 🛠 The hardware challenges in robotics include creating reliable, lightweight, and affordable robotic hands.
- 💡 The software side of robotics is identified as particularly challenging due to uncertainties in control, perception, and physics.
- 👀 Advances like LIDAR have improved 3D perception for robots, but they still have limitations, especially with shiny or transparent objects.
- 📦 E-commerce has become a 'sweet spot' for robotics, with companies like Ambi Robotics using AI and deep learning to improve package handling.
- 🍽 The speaker discusses ongoing research into making robots more capable in homes, including handling deformable objects and performing tasks like untangling knots and folding laundry.
- 🚀 Despite the progress, the speaker acknowledges that we're not yet at the level of home robots seen in futuristic visions, but encourages patience as development continues.
Q & A
What is Moravec's paradox as mentioned in the script?
-Moravec's paradox refers to the phenomenon that what is easy for robots, such as lifting heavy objects, is difficult for humans, and what is easy for humans, like manipulating small or delicate objects, is very hard for robots. This paradox highlights the persistent challenge in robotics of grasping arbitrary objects.
Why are simple hands like the parallel jaw gripper and suction cup preferred in robotics?
-Simple hands are favored in robotics due to their reliability, lightweight design, and cost-effectiveness. They are also easier to maintain and can be effective for a wide range of tasks, as demonstrated by the video where simple grippers perform complex actions.
How does LIDAR technology assist in addressing the perception challenges for robots?
-LIDAR, which uses light beams to create a three-dimensional model of the environment, helps robots perceive their surroundings more accurately. It provides a breakthrough in understanding the spatial layout of objects, although it still has limitations with shiny or transparent objects that can cause unpredictable light reflections.
What is the significance of the Dex-net system in the context of the script?
-Dex-net is a system developed by the speaker and their students that enables robots to reliably pick up objects. It uses AI and deep learning to train the robot in simulation, allowing it to grasp a variety of objects it has not been specifically trained on, which is a significant advancement in the field of robotics.
What is the current application of the technology developed by the speaker's company, Ambi Robotics?
-Ambi Robotics applies the technology developed at UC Berkeley to sort packages in the e-commerce industry. Their machines pick up, scan, and sort packages into smaller bins based on zip codes, addressing the high demand for efficient package delivery.
Why is manipulating deformable objects considered a challenge for robots?
-Manipulating deformable objects is challenging for robots because these objects can change shape and are often difficult to grasp and control. Robots need to understand the physical properties and behavior of these objects to interact with them effectively, which is a complex problem in robotics.
What is the 'two-second fold' technique mentioned in the script, and how does it relate to robot folding laundry?
-The 'two-second fold' is a technique used to demonstrate the efficiency of a robot folding laundry. The robot performs the same folding action as a human but in a slightly longer time frame, which is still considered real-time and not sped up, showcasing progress in the speed of robotic folding.
How does the speaker's research on untangling knots and folding laundry illustrate the progress and challenges in robotics?
-The speaker's research on untangling knots and folding laundry demonstrates the progress in robotics by showing that robots can now perform tasks that were once thought to be extremely difficult. However, it also highlights the challenges, as these tasks are still not performed at the speed or reliability that would be practical for everyday use.
What is the role of AI and deep learning in the development of robots as discussed in the script?
-AI and deep learning play a crucial role in the development of robots by enabling them to learn and improve their grasping abilities through simulation. This technology allows robots to 'dream' about how to grasp objects and learn to do so reliably, which is a significant step towards more autonomous and capable robots.
Why is the speaker advocating for patience when it comes to the development of home robots?
-The speaker advocates for patience because, despite significant advancements, there is still a gap between the capabilities of robots seen in popular culture and the reality of current technology. Developing robots that can perform complex tasks in unstructured environments like homes is a challenging process that requires time and continued research.
Outlines
🤖 The Gap Between Fiction and Reality in Robotics
The speaker begins by expressing the common desire for helpful robots at home and humorously questions the delay in their widespread availability despite advancements in technology. They introduce themselves as a researcher with 30 years of experience at UC Berkeley and set the stage for a discussion on the disparity between the fictional portrayal of robots and their real-world capabilities. The speaker highlights Moravec's paradox, which points out that tasks easy for robots, like lifting heavy objects, are hard for humans, and vice versa. They emphasize the ongoing challenge of robots grasping arbitrary objects, a problem that has persisted in the field. The speaker also shares personal anecdotes about their clumsiness as a child and their lifelong dedication to making robots less awkward. They delve into the intricacies of robotic hardware, advocating for simple designs like the parallel jaw gripper and suction cups over complex, unreliable, and expensive alternatives.
🛠️ Tackling the Challenges of Robotic Grasping in E-commerce
The speaker transitions into discussing the software side of robotics, where uncertainty in control, perception, and physics poses significant challenges. They explain the difficulties in precise control due to mechanical imprecision and the limitations of sensors, including high-resolution cameras and LIDAR, which struggle with shiny or transparent objects. The introduction of tactile sensors is mentioned as a potential advancement, though it's still in early stages. The unpredictability of physics in robotic manipulation is illustrated through an example of a robot pushing a bottle, which can end up in different places due to microscopic surface variations. The speaker then shifts focus to e-commerce, identifying it as a niche where robots could excel due to the pandemic-driven boom in online shopping. They describe the current manual nature of warehouse order fulfillment and the high turnover due to the laborious task. The speaker shares their own work on a system called Dex-net, which uses AI and deep learning to teach robots to grasp objects reliably. This technology has been commercialized through their company, Ambi Robotics, which helps sort packages in warehouses. While progress is being made, the speaker acknowledges that this is not the home robot many envision.
🔍 Advancing Robotics for Everyday Tasks
In the final paragraph, the speaker discusses ongoing research aimed at making robots more capable for home use, focusing on handling deformable objects. They describe a project to untangle knots using a robot that analyzes and manipulates cables, achieving an 80% success rate. The speaker also addresses the long-standing challenge of robotic laundry folding, a task that has been notoriously slow. They share their team's efforts to develop a two-armed robot that can fold laundry faster by mimicking human actions like flinging and smoothing. Another project involves teaching robots to bag items, a task complicated by the variable configurations of bags. Using fluorescent paint and lights, the robot learns to manipulate bags, achieving a 50% success rate. The speaker concludes by reiterating Moravec's paradox and acknowledging the public's patience for the Jetsons-like robots that are still not a reality. They express optimism for the future of robotics, emphasizing the mutual need between humans and robots, with the latter requiring human assistance for tasks they cannot yet perform.
Mindmap
Keywords
💡Robotics
💡Moravec's Paradox
💡Gripper
💡Uncertainty
💡LIDAR
💡Tactile Sensor
💡E-commerce
💡Dex-Net
💡Deformable Objects
💡Ambi Robotics
💡Robotic Manipulation
Highlights
The audience is eager for home robots but they are not yet common.
Existing home robots are not performing exciting tasks.
The speaker has been researching robots at UC Berkeley for 30 years.
Moravec's paradox: Tasks easy for robots are hard for humans, and vice versa.
Grasp arbitrary objects is a significant challenge in robotics.
The speaker's personal journey from being clumsy to a career in robotics.
The complexity and unreliability of robot hands with many motors and tendons.
Advocacy for simple robot hands like the parallel jaw gripper.
Suction cups as an even simpler robot gripper used in industry.
Uncertainty in robotics due to control, perception, and physics.
LIDAR as a breakthrough in robotic perception but not a perfect solution.
Tactile sensors are in early stages of development for robotics.
Physics of object manipulation is unpredictable for robots.
E-commerce is a 'sweet spot' for robots in package handling.
Dex-net, a system developed by the speaker's team, reliably picks up objects.
Ambi Robotics, a company founded by the speaker, uses AI for package sorting.
New research aims to make robots capable of handling deformable objects.
A project to untangle knots using robots has achieved an 80% success rate.
Roboticists are working on speeding up the process of folding laundry.
Bagging is a challenging task for robots due to the variability of bag configurations.
Robots are still clumsy but progress is being made in their capabilities.
The speaker concludes by emphasizing the need for patience as robots continue to evolve.
Transcripts
I have a feeling most people in this room would like to have a robot at home.
It'd be nice to be able to do the chores and take care of things.
Where are these robots?
What's taking so long?
I mean, we have our tricorders,
and we have satellites.
We have laser beams.
But where are the robots?
(Laughter)
I mean, OK, wait, we do have some robots in our home,
but, not really doing anything that exciting, OK?
(Laughter)
Now I've been doing research at UC Berkeley for 30 years
with my students on robots,
and in the next 10 minutes,
I'm going to try to explain the gap between fiction and reality.
Now we’ve seen images like this, right?
These are real robots.
They're pretty amazing.
But those of us who work in the field,
well, the reality is more like this.
(Laughter)
That's 99 out of 100 times, that's what happens.
And in the field, there's something that explains this
that we call Moravec's paradox.
And that is, what's easy for robots,
like being able to pick up a large object,
large, heavy object,
is hard for humans.
But what's easy for humans,
like being able to pick up some blocks and stack them,
well, it turns out that is very hard for robots.
And this is a persistent problem.
So the ability to grasp arbitrary objects is a grand challenge for my field.
Now by the way, I was a very klutzy kid.
(Laughter)
I would drop things.
Any time someone would throw me a ball, I would drop it.
I was the last kid to get picked on a basketball team.
I'm still pretty klutzy, actually,
but I have spent my entire career studying how to make robots less clumsy.
Now let's start with the hardware.
So the hands.
Now this is a robot hand, a particular type of hand.
It's a lot like our hand.
And it has a lot of motors, a lot of tendons
and cables as you can see.
So it's unfortunately not very reliable.
It's also very heavy and very expensive.
So I'm in favor of very simple hands, like this.
So this has just two fingers.
It's known as a parallel jaw gripper.
So it's very simple.
It's lightweight and reliable and it's very inexpensive.
And if you're doubting that simple hands can be effective,
look at this video where you can see that two very simple grippers,
these are being operated, by the way,
by humans who are controlling the grippers like a puppet.
But very simple grippers are capable of doing very complex things.
Now actually in industry,
there’s even a simpler robot gripper, and that’s the suction cup.
And that only makes a single point of contact.
So again, simplicity is very helpful in our field.
Now let's talk about the software.
This is where it gets really, really difficult
because of a fundamental issue, which is uncertainty.
There's uncertainty in the control.
There’s uncertainty in the perception.
And there’s uncertainty in the physics.
Now what do I mean by the control?
Well if you look at a robot’s gripper trying to do something,
there's a lot of uncertainty in the cables and the mechanisms
that cause very small errors.
And these can accumulate and make it very difficult to manipulate things.
Now in terms of the sensors, yes,
robots have very high-resolution cameras just like we do,
and that allows them to take images of scenes in traffic
or in a retirement center,
or in a warehouse or in an operating room.
But these don't give you the three-dimensional structure
of what's going on.
So recently, there was a new development called LIDAR,
and this is a new class of cameras that use light beams to build up
a three-dimensional model of the environment.
And these are fairly effective.
They really were a breakthrough in our field, but they're not perfect.
So if the objects have anything that's shiny or transparent,
well, then the light acts in unpredictable ways,
and it ends up with noise and holes in the images.
So these aren't really the silver bullet.
And there’s one other form of sensor out there now called a “tactile sensor.”
And these are very interesting.
They use cameras to actually image the surfaces
as a robot would make contact,
but these are still in their infancy.
Now the last issue is the physics.
And let me illustrate for you by showing you,
we take a bottle on a table
and we just push it,
and the robot's pushing it in exactly the same way each time.
But you can see that the bottle ends up in a very different place each time.
And why is that?
Well it’s because it depends on the microscopic surface topography
underneath the bottle as it slid.
For example, if you put a grain of sand under there,
it would react very differently than if there weren't a grain of sand.
And we can't see if there's a grain of sand because it's under the bottle.
It turns out that we can predict the motion of an asteroid
a million miles away,
far better than we can predict the motion of an object
as it's being grasped by a robot.
Now let me give you an example.
Put yourself here into the position of being a robot.
You're trying to clear the table
and your sensors are noisy and imprecise.
Your actuators, your cables and motors are uncertain,
so you can't fully control your own gripper.
And there's uncertainty in the physics,
so you really don't know what's going to happen.
So it's not surprising that robots are still very clumsy.
Now there's one sweet spot for robots, and that has to do with e-commerce.
And this has been growing, it's a huge trend.
And during the pandemic, it really jumped up.
I think most of us can relate to that.
We started ordering things like never before,
and this trend is continuing.
And the challenge is to meet the demand,
we have to be able to get all these packages delivered in a timely manner.
And the challenge is that every package is different,
every order is different.
So you might order some some nail polish and an electric screwdriver.
And those two objects are going to be
somewhere inside one of these giant warehouses.
And what needs to be done is someone has to go in,
find the nail polish and then go and find the screwdriver,
bring them together, put them into a box and deliver them to you.
So this is extremely difficult, and it requires grasping.
So today, this is almost entirely done with humans.
And the humans don't like doing this work,
there's a huge amount of turnover.
So it's a challenge.
And people have tried to put robots
into warehouses to do this work.
(Laughter)
It hasn't turned out all that well.
But my students and I, about five years ago,
we came up with a method, using advances in AI and deep learning,
to have a robot essentially train itself to be able to grasp objects.
And the idea was that the robot would do this in simulation.
It was almost as if the robot were dreaming about how to grasp things
and learning how to grasp them reliably.
And here's the result.
This is a system called Dex-net
that is able to reliably pick up objects
that we put into these bins in front of the robot.
These are objects it's never been trained on,
and it's able to pick these objects up
and reliably clear these bins over and over again.
So we were very excited about this result.
And the students and I went out to form a company,
and we now have a company called Ambi Robotics.
And what we do is make machines that use the algorithms,
the software we developed at Berkeley,
to pick up packages.
And this is for e-commerce.
The packages arrive in large bins, all different shapes and sizes,
and they have to be picked up,
scanned and then put into smaller bins depending on their zip code.
We now have 80 of these machines operating across the United States,
sorting over a million packages a week.
Now that’s some progress,
but it's not exactly the home robot that we've all been waiting for.
So I want to give you a little bit of an idea
of some the new research that we're doing
to try to be able to have robots more capable in homes.
And one particular challenge is being able to manipulate deformable objects,
like strings in one dimension,
two-dimensional sheets and three dimensions,
like fruits and vegetables.
So we've been working on a project to untangle knots.
And what we do is we take a cable and we put that in front of the robot.
It has to use a camera to look down, analyze the cable,
figure out where to grasp it
and how to pull it apart to be able to untangle it.
And this is a very hard problem,
because the cable is much longer than the reach of the robot.
So it has to go through and manipulate, manage the slack as it's working.
And I would say this is doing pretty well.
It's gotten up to about 80 percent success
when we give it a tangled cable at being able to untangle it.
The other one is something I think we also all are waiting for:
robot to fold the laundry.
Now roboticists have actually been looking at this for a long time,
and there was some research that was done on this.
But the problem is that it's very, very slow.
So this was about three to six folds per hour.
(Laughter)
So we decided to to revisit this problem
and try to have a robot work very fast.
So one of the things we did was try to think
about a two-armed robot that could fling the fabric
the way we do when we're folding,
and then we also used friction in this case to drag the fabric
to smooth out some wrinkles.
And then we borrowed a trick which is known as the two-second fold.
You might have heard of this.
It's amazing because the robot is doing exactly the same thing
and it's a little bit longer, but that's real time,
it's not sped up.
So we're making some progress there.
And the last example is bagging.
So you all encounter this all the time.
You go to a corner store, and you have to put something in a bag.
Now it's easy, again, for humans,
but it's actually very, very tricky for robots
because for humans, you know how to take the bag
and how to manipulate it.
But robots, the bag can arrive in many different configurations.
It’s very hard to tell what’s going on
and for the robot to figure out how to open up that bag.
So what we did was we had the robot train itself.
We painted one of these bags with fluorescent paint,
and we had fluorescent lights that would turn on and off,
and the robot would essentially teach itself how to manipulate these bags.
And so we’ve gotten it now up to the point
where we're able to solve this problem about half the time.
So it works,
but I'm saying, we're still not quite there yet.
So I want to come back to Moravec's paradox.
What's easy for robots is hard for humans.
And what's easy for us is still hard for robots.
We have incredible capabilities.
We're very good at manipulation.
(Laughter)
But robots still are not.
I want to say, I understand.
It’s been 60 years,
and we're still waiting for the robots that the Jetsons had.
Why is this difficult?
We need robots because we want them to be able to do tasks that we can't do
or we don't really want to do.
But I want you to keep in mind that these robots, they're coming.
Just be patient.
Because we want the robots,
but robots also need us
to do the many things that robots still can't do.
Thank you.
(Applause)
Ver Más Videos Relacionados
5.0 / 5 (0 votes)