Elon Musk has finally unveiled a prototype of the much-hyped Optimus robot — a bipedal machine that the Tesla boss envisions will one day be sold as a “general purpose”[ads1]; robot cheaper than a car and just as capable of work in factories and doing tasks at home.
The presentation of Optimus at Tesla’s AI Day was peppered with the usual Musk rushes of hype and misdirection. But it also gave roboticists an admirably detailed look at some of the machine’s actual capabilities. Their judgment? That Tesla’s engineers have done incredible work in such a short time, but that Optimus is still more hype than reality: a research robot that will take years to develop into something meaningful.
Before we dive into the experts’ opinions, however, let’s recap what we saw on the day.
- Two robots were shown. The first, called Bumble C, was a “rough development” robot that was seen walking across the stage and performing tasks in pre-recorded demos such as picking up a watering can and moving boxes. The second robot was described by Musk as “pretty close to what will go into production” and had an external cover. But it was just used as a prop, rolling around behind Musk and just showing waving hands.
- Musk reiterated his ambition to create “a useful humanoid robot as quickly as possible,” one that could be “made in very high volume, eventually millions of units” and that would “cost a lot less than a car — a lot less than $20,000 ». He later claimed that inventing such a machine would provide “a future in which there is no poverty” and a “fundamental transformation of civilization as we know it.”
- We have some basic specs for Optimus (though, as with the concept robot, it’s unclear if they’ll be true the next time we get an update). It weighs 73 kg, has 28 degrees of freedom in its limbs, and is powered by a 2.3 kWh battery that Tesla claims will last an entire working day. The excellent mechanical function is hand-modeled after that of a human, and Tesla showed how it has transferred the software used to power self-driving cars for navigation and control.
- In a Q&A at the end of the presentation, Musk echoed speculation that the robot could “be like a friend” one day, saying that customers will be able to order one “within three years, probably no more than five years” (although it’s not it is clear whether he meant companies or individuals).
So, it’s just the facts of what was said and shown on stage. But what did the experts have to say? Here are some of their reactions, from interviews with The Verge and judgments shared on social media.
“Am I impressed? No. Am I laughing? No.”
Christian Hubicki, professor of robotics at Florida State University, offered their thoughts in a Twitter threadand notes, like many others, that the Tesla team “came a long way in about a year” but that the characteristics of Optimus actually shown on stage “seem standard (but not sensational) for humanoids.”
In his thread, Hubicki notes that the Optimus robot appears to use a walking method known in robotics as zero moment point, or ZMP (although this has not been confirmed by Tesla). This is a movement method that has been in use for decades, deployed in well-known robots such as Honda’s Asimo. (Looking at Optimus, you can probably see the similarity in how the knees are bent in a half-crouch, and it carefully shifts its weight from one foot to the other.) Hubicki notes that this method of walking is “quite safe, but not mind-blowing in 2022.”
He also points out that another big question not answered by the presentation is reliability. “How often does it fall down? You can’t tell from a cool video – or even a live demo, he says. When it comes to putting robots in factories, reliability is obviously a huge factor, as downtime anywhere on an assembly line can have a significant ripple effect.
“An extraordinarily brave demonstration of a Herculean effort that unfortunately lacks novelty and imagination”
Will Jackson, CEO of Engineered Arts, said The Verge the first two prototypes of the Optimus bot were “deliberately” missing with “no novelty” in the design.
“They are very similar in concept to Honda’s Asimo robots, whose development has now been abandoned,” Jackson said. “The overall design is heavily built, clunky and power-inefficient – the hands are very basic, the only redeeming feature being a clutch mechanism in the finger actuation. If you want to know how far they are from human-level movement and capabilities, compare last year’s reveal – a man in a robot suit – with the actual hardware of the year.”
Jackson praised Tesla’s AI work in general and the dedication of the company’s engineers, but questioned the very idea of building a humanoid robot for poor labor in the first place. “I’m amazed that Musk can address an audience so enamored with the idea of a humanoid and completely fail to realize that their desire to interact with a robot is the killer application. Did he think they were applauding because the world will finally have a humanoid robot that can lift a pipe in a car factory?”
He concluded that the demo showed “an exceptionally brave live demonstration of a Herculean effort that sadly lacks novelty and imagination” and that “hopefully we’ll see a course correction by next year’s event.”
An impressive display, but many unanswered questions
Jonathan Aitken, a roboticist and teacher at the University of Sheffield in the UK, told The Verge that it was an “interesting demo” and that the “biggest thing is the progress they’ve made over the time they’ve been working on things.” But he also noted that there were many unanswered questions about the robot’s capabilities and the exact nature of the demos.
There is always a big difference between pre-recorded demos and actual work
Aitken points out that the pre-recorded videos of Optimus showed an “umbilical cord” attached to the robot, for example. “That to me poses two big questions, power and communication / control.” Was the wire there for safety? Or for electricity and instructions? (It was noted on stage that it was the first time the robot walked without a tether.)
“From the presentation, it was good to see the range of activities that the robot was working on,” says Aitken. “Although they were relatively limited.” He notes that footage of the robot’s moving components in a Tesla factory workstation was impressive, but lacked context to explain how meaningful the work was. “I would like to see more information about how big a component is and how general this is in their manufacturing process.”
“Are you releasing features for the sake of features or to actually solve problems?”
Cynthia Yeung, product manager at robot logistics company Plus One Robotics, live-tweeted her thoughts on the presentationquestions the company’s focus on replicating the human form so closely and its planned future business model (which Musk says will involve selling the robots directly rather than selling their services as many robot startups do).
“I think @elonmusk seems to be in love with the @BostonDynamics approach to robots (form over function) as opposed to what many others are working on (function informs form),” Yeung tweeted.
Yeung noted that the presentation focused a lot on how the team had made Optimus go, but that solving this challenge would not necessarily solve real problems. She also pointed out that the robot’s five-fingered hand was not necessarily as good as simpler two- or three-fingered pinchers or vacuum systems. “There’s a reason all warehouse startups don’t use hand-like manipulation mechanisms,” Yeung tweeted.
Elsewhere, Yeung praised Musk to share the spotlight with the team of engineers and said the company made some “cool” actuators and simulators. But her overall assessment of the technical achievements was damning. “None of this is innovative. Hire some PhDs and go to some robotics conferences,” Yeung tweeted.
“So far they’ve built a nice platform to do the research.”
Talking to The Verge via email, University of California San Diego robotics professor Henrik Christensen said the mechanical design and walking skills of Optimus were “solid” but that “there is little innovation” in Tesla’s approach.
“There’s no real evidence that it can do basic navigation, grasping, manipulation.”
Like Yeung, Christensen points out that Optimus’ five-fingered hand design was probably unnecessary (“not sure you need 5 fingers … I didn’t see a solid argument here beyond hand-waving”). And he notes that a lot was simply missing from Tesla’s presentation: “There’s no real evidence that it can do basic navigation, grip, manipulation.”
“It’s a good initial design and it’s impressive how far they’ve come in 9 months, but the innovation beyond Boston Dynamics and Agility is very limited,” Christensen said. “So far they’ve built a nice platform to do the research […] The good news is that Musk is not afraid to think big and invest. As such, I expect to see real innovation here in the future.”
What Musk can really offer: attention and funding
Animesh Gargan assistant professor of AI at the University of Toronto, offered a response that focused heavily on the broader challenges facing the field—that is, why develop a general-purpose humanoid robot when simpler systems are more achievable?
Writing in a post shared on Google Docs, Garg notes that creating humanoid robots is “a zero-to-one problem with no revenue until the system works to deliver value.” That makes the investment challenging. But he compares the approach to designing fully autonomous cars, noting that if the end goal can be reached (seamless integration with existing human environments), then we save a lot by not having to restructure infrastructure to accommodate machines (be they robots or cars) ).
Garg has some slight criticisms of Optimus’ design, noting, for example, that the cable-operated hands seem to have a decent load capacity but a slow response time. But he is more positive that Musk and Tesla will be involved in this space at all:
Overall, the current design is a very good first step. Interest in building such systems is welcome because Tesla and Elon Musk’s involvement in the problem brings attention, talent and resources to the problem, setting in motion a flywheel of progress. This effort should be praised with cautious optimism by the community, because the compass is pointing in the right direction, and Elon is carrying the weight of Tesla engineers as we go through the AI/Robotics jungle.