Let’s talk about humanoids for a minute, right? Why do so many roboticists insist on making robots like us? Get ready to see many more humanoid robots.
This topic is relevant for several reasons. First, and most notably, is the fact that Tesla is planning to introduce a version of Optimus (aka Tesla Bot) that won’t just be a man in spandex. Tesla describes the project this way:
Develop the next generation of automation, including a bipedal, general-purpose humanoid robot capable of performing unsafe, repetitive, or boring tasks. We are looking for Mechanical, Electrical, Control Systems and Software Engineers to help us leverage our AI expertise beyond our fleet.
Musk, who announced that a prototype could be ready as early as next year, has faced criticism from the robotics community for an ambitious, if not impossible, project. The robot’s upcoming debut will have to do more than just take the stage to silence the naysayers. The biggest question that arises over all of this is: if it’s not impossible, why didn’t some smart and well-funded minds make it? It’s certainly not for lack of trying.
Another reason why this is the main thing is due to our story that happened this morning, which broke the news about Figure’s self-funded effort to represent its own humanoid thanks to an impressive staff of former employees of Apple, Tesla, Boston Dynamics and Google. Given that the project hasn’t yet unveiled a product — or, for that matter, a company — that means it’s still too early to judge the proposal. Of course, the above question remains.
A little easier to answer the question, why a humanoid? This is something I have discussed with a number of roboticists over the years. Our brains are wired to think of robots as mechanical versions of ourselves. Decades of science fiction have taken care of that. But the robotics approach is — more often than not — pragmatic. The right form factor for the job is a good rule of thumb. By going beyond that, you introduce more potential points of failure while driving up the price.
There’s a reason the world’s most popular consumer robot is the hockey puck vacuum cleaner. It is designed to do one particular job well (improve it over generations) in the form most suited to that job. Incorporating some human traits (a la Amazon’s Astro) would help impersonate the robot and perhaps allow users to form an emotional connection with the thing, but it doesn’t have to. And iRobot is already hard enough delivering MSRP below $1,000.
However, the counterargument is persuasive in itself. A few years ago, I spoke to some of the team that tested NASA’s Valkyrie bipedal robot. As they note, people tend to shape their world around them. We build buildings and streets to our own evolutionary specifications, so it follows that a robot designed to navigate these spaces will end up looking like us. Automation is the most sincere form of flattery.
We will keep a close eye on this space.
Speaking of humanoids, it seems that SoftBank Robotics Europe’s journey has finally come to an end. After the acquisition in 2015, Aldebaran eventually ran into trouble switching from the NAO research robot to Pepper. The latter was built on the premise that a friendly face built on top of a robot with limited functionality could help drive traffic for a business.
Following reports that Pepper was being discontinued over the summer, SoftBank sold the company to German firm United Robotics Group. This week, URG announced that it will return the brand to its original name while working to “improve our offerings for existing products like Pepper and Nao.” Meanwhile, SoftBank remains a shareholder.
Meanwhile, Nvidia this week uploaded a whole bunch of product news, including several articles regarding his efforts to build a robotic platform. In particular, CEO Jensen Huang detailed the chipmaker’s push to move its Isaac Sim robotics simulator to the cloud via AWS RoboMaker. NVIDIA notes:
Using Isaac Sim in the cloud, roboticists will be able to generate large datasets from physically accurate sensor simulations to train AI-based perceptual models on their robots. The synthetic data generated from these simulations improves model performance and provides training data that is often impossible to collect in the real world.
More news from the “general purpose” robotics front Apptronik discusses his future robot Apollo. The Austin firm has already closed a deal to bring their own humanoid to NASA.
“Traditional robots are really designed to perform very repetitive actions in structured environments,” co-founder and CEO Jeff Cardenas tells TechCrunch. “What we’ve been really focused on is how do we build robots that can operate in highly variable dynamic environments? With a humanoid robot, really, how can we build a robot built by people for people to work in places meant for people?”
Not surprisingly, the company positions the system as a platform on which developers can create a set of different features. Apptronik says it hopes to showcase the Apollo at next year’s SXSW in its hometown of Austin.
More big news about tiny robots this week, as Devin reveals Ant robots at Cornell University, which are “actually ant-to-ant-sized” if you can imagine such a thing. The systems use a photovoltaic cell to power it and microscopic circuits to move its tiny legs. Potential applications are the standard variety listed for this type of very small robot.
Applications will range from environmental cleaning and monitoring to targeted drug delivery, cell monitoring or stimulation, and microscopic surgery. In all these applications, robots with built-in controls for environmental detection and response and autonomous operation provide a distinct advantage, creating the basis for ubiquitous intelligent microscopic robots that can create positive results in the world around us.
A smaller but rather funny robot from KEYi Tech which, as I write this, ready to cross the $1 million mark on Kickstarter, more than a month left before the release. Comparisons between Loona and Anki’s Cozmo robot are inevitable, but ClicBot’s creator has done a truly incredible job on the robot’s facial expressions and movements. As I mentioned in my Anki article, I asked the company to send me a raw video to confirm it was not a render.
Finally, over the week, $5 million in funding will go to Civil robotics. A San Francisco-based firm has created an autonomous robot to create land surveys for construction sites. Co-founder and CEO Tom Yeshurun says:
The construction industry is facing labor shortages and CivDot is improving efficiency and safety in the workplace while moving projects forward from the start. Bechtel, a leader in the EPC industry, among many others, has already adopted CivDots for filming. Today’s funding demonstrates the opportunity before us as a company to build the world around us.
The seed round was hosted by ff Venture Capital and Alley Robotics Ventures with participation from Trimble Ventures. So many adventures, so little time.
Let’s talk about robots. Subscribe to Actuator.
ShareShare article via FacebookShare article via TwitterShare article via LinkedInShare article via email Exchange Ted…
Residents of the temporarily occupied Crimea record the activity of enemy aircraft in the various…
Photo: TNT station September 26 at 20:00 on TNT begins a new comedy with the…
Benefits for mobilized Russians for notarial services were provided by the Federal Chamber of NotariesA…
Madrid (EFE). Leading feminist organizations have drafted about twenty "amendments" to the transgender bill, which…
Transfer work in Beşiktaş started early. In the black and white team, coach Valerien Ismael…