Stanford’s Doggo is a petite robotic quadruped you can (maybe) build yourself

Got a few thousand bucks and a good deal of engineering expertise? You’re in luck: Stanford students have created a quadrupedal robot platform called Doggo that you can build with off-the-shelf parts and a considerable amount of elbow grease. That’s better than the alternatives, which generally require a hundred grand and a government-sponsored lab.

Due to be presented (paper on arXiv here) at the IEEE International Conference on Robots and Automation, Doggo is the result of research by the Stanford Robotics Club, specifically the Extreme Mobility team. The idea was to make a modern quadrupedal platform that others could build and test on, but keep costs and custom parts to a minimum.

The result is a cute little bot with rigid-looking but surprisingly compliant polygonal legs that has a jaunty, bouncy little walk and can leap more than three feet in the air. There are no physical springs or shocks involved, but by sampling the forces on the legs 8,000 times per second and responding as quickly, the motors can act like virtual springs.

It’s limited in its autonomy, but that’s because it’s built to move, not to see and understand the world around it. That is, however, something you, dear reader, could work on. Because it’s relatively cheap and doesn’t involve some exotic motor or proprietary parts, it could be a good basis for research at other robotics departments. You can see the designs and parts necessary to build your own Doggo right here.

“We had seen these other quadruped robots used in research, but they weren’t something that you could bring into your own lab and use for your own projects,” said Doggo lead Nathan Kau in a Stanford news post. “We wanted Stanford Doggo to be this open source robot that you could build yourself on a relatively small budget.”

In the meantime the Extreme Mobility team will be both improving on the capabilities of Doggo by collaborating with the university’s Robotic Exploration Lab, and also working on a similar robot but twice the size — Woofer.

Massive 2015 natural gas leak caused by microbial corrosion, report says

Aliso Canyon methane leak 6.

A massive natural gas leak at a storage facility in Southern California was caused by microbial corrosion of well equipment, according to a new independent report from analysis firm Blade Energy Partners. The report blames the storage facility owner, Southern California Gas (SoCalGas) for failing to conduct follow-up inspections of equipment, despite knowing about 60 smaller leaks at the facility that had occurred since the 1970s.

The final leak—which spewed 109,000 metric tons of methane into the air over five months between 2015 and 2016—was the biggest methane leak in US history. (A larger loss of methane occurred in 2004 in Texas, but a corresponding fire immediately combusted the methane into carbon dioxide.) But the California leak at the Aliso Canyon Natural Gas Storage Field was particularly devastating because methane, unfortunately, is a far more potent greenhouse gas than carbon dioxide.

The new report (PDF) was commissioned three years earlier to find the root cause of the leak. According to a press release from the California Public Utilities Commission (CPUC), Blade Energy Partners found that the leak came from a seven-inch outer well casing which had corroded due to exposure to microbes from groundwater. The natural gas storage facility at Aliso Canyon is made up of dozens of vast underground caverns which were previously filled with oil before they were pumped and emptied decades ago. Since then, the caverns have been used to store natural gas to supply the Southern California area.

Read 5 remaining paragraphs | Comments

Why is Facebook doing robotics research?

It’s a bit strange to hear that the world’s leading social network is pursuing research in robotics rather than, say, making search useful, but Facebook is a big organization with many competing priorities. And while these robots aren’t directly going to affect your Facebook experience, what the company learns from them could be impactful in surprising ways.

Though robotics is a new area of research for Facebook, its reliance on and bleeding-edge work in AI are well known. Mechanisms that could be called AI (the definition is quite hazy) govern all sorts of things, from camera effects to automated moderation of restricted content.

AI and robotics are naturally overlapping magisteria — it’s why we have an event covering both — and advances in one often do the same, or open new areas of inquiry, in the other. So really it’s no surprise that Facebook, with its strong interest in using AI for a variety of tasks in the real and social media worlds, might want to dabble in robotics to mine for insights.

What then could be the possible wider applications of the robotics projects it announced today? Let’s take a look.

Learning to walk from scratch

“Daisy” the hexapod robot.

Walking is a surprisingly complex action, or series of actions, especially when you’ve got six legs, like the robot used in this experiment. You can program in how it should move its legs to go forward, turn around, and so on, but doesn’t that feel a bit like cheating? After all, we had to learn on our own, with no instruction manual or settings to import. So the team looked into having the robot teach itself to walk.

This isn’t a new type of research — lots of roboticists and AI researchers are into it. Evolutionary algorithms (different but related) go back a long way, and we’ve already seen interesting papers like this one:

By giving their robot some basic priorities like being “rewarded” for moving forward, but no real clue how to work its legs, the team let it experiment and try out different things, slowly learning and refining the model by which it moves. The goal is to reduce the amount of time it takes for the robot to go from zero to reliable locomotion from weeks to hours.

What could this be used for? Facebook is a vast wilderness of data, complex and dubiously structured. Learning to navigate a network of data is of course very different from learning to navigate an office — but the idea of a system teaching itself the basics on a short timescale given some simple rules and goals is shared.

Learning how AI systems teach themselves, and how to remove roadblocks like mistaken priorities, cheating the rules, weird data-hoarding habits and other stuff is important for agents meant to be set loose in both real and virtual worlds. Perhaps the next time there is a humanitarian crisis that Facebook needs to monitor on its platform, the AI model that helps do so will be informed by the autodidactic efficiencies that turn up here.

Leveraging “curiosity”

Researcher Akshara Rai adjusts a robot arm in the robotics AI lab in Menlo Park. (Facebook)

This work is a little less visual, but more relatable. After all, everyone feels curiosity to a certain degree, and while we understand that sometimes it kills the cat, most times it’s a drive that leads us to learn more effectively. Facebook applied the concept of curiosity to a robot arm being asked to perform various ordinary tasks.

Now, it may seem odd that they could imbue a robot arm with “curiosity,” but what’s meant by that term in this context is simply that the AI in charge of the arm — whether it’s seeing or deciding how to grip, or how fast to move — is given motivation to reduce uncertainty about that action.

That could mean lots of things — perhaps twisting the camera a little while identifying an object gives it a little bit of a better view, improving its confidence in identifying it. Maybe it looks at the target area first to double check the distance and make sure there’s no obstacle. Whatever the case, giving the AI latitude to find actions that increase confidence could eventually let it complete tasks faster, even though at the beginning it may be slowed by the “curious” acts.

What could this be used for? Facebook is big on computer vision, as we’ve seen both in its camera and image work and in devices like Portal, which (some would say creepily) follows you around the room with its “face.” Learning about the environment is critical for both these applications and for any others that require context about what they’re seeing or sensing in order to function.

Any camera operating in an app or device like those from Facebook is constantly analyzing the images it sees for usable information. When a face enters the frame, that’s the cue for a dozen new algorithms to spin up and start working. If someone holds up an object, does it have text? Does it need to be translated? Is there a QR code? What about the background, how far away is it? If the user is applying AR effects or filters, where does the face or hair stop and the trees behind begin?

If the camera, or gadget, or robot, left these tasks to be accomplished “just in time,” they will produce CPU usage spikes, visible latency in the image, and all kinds of stuff the user or system engineer doesn’t want. But if it’s doing it all the time, that’s just as bad. If instead the AI agent is exerting curiosity to check these things when it senses too much uncertainty about the scene, that’s a happy medium. This is just one way it could be used, but given Facebook’s priorities it seems like an important one.

Seeing by touching

Although vision is important, it’s not the only way that we, or robots, perceive the world. Many robots are equipped with sensors for motion, sound, and other modalities, but actual touch is relatively rare. Chalk it up to a lack of good tactile interfaces (though we’re getting there). Nevertheless, Facebook’s researchers wanted to look into the possibility of using tactile data as a surrogate for visual data.

If you think about it, that’s perfectly normal — people with visual impairments use touch to navigate their surroundings or acquire fine details about objects. It’s not exactly that they’re “seeing” via touch, but there’s a meaningful overlap between the concepts. So Facebook’s researchers deployed an AI model that decides what actions to take based on video, but instead of actual video data, fed it high-resolution touch data.

Turns out the algorithm doesn’t really care whether it’s looking at an image of the world as we’d see it or not — as long as the data is presented visually, for instance as a map of pressure on a tactile sensor, it can be analyzed for patterns just like a photographic image.

What could this be used for? It’s doubtful Facebook is super interested in reaching out and touching its users. But this isn’t just about touch — it’s about applying learning across modalities.

Think about how, if you were presented with two distinct objects for the first time, it would be trivial to tell them apart with your eyes closed, by touch alone. Why can you do that? Because when you see something, you don’t just understand what it looks like, you develop an internal model representing it that encompasses multiple senses and perspectives.

Similarly, an AI agent may need to transfer its learning from one domain to another — auditory data telling a grip sensor how hard to hold an object, or visual data telling the microphone how to separate voices. The real world is a complicated place and data is noisier here — but voluminous. Being able to leverage that data regardless of its type is important to reliably being able to understand and interact with reality.

So you see that while this research is interesting in its own right, and can in fact be explained on that simpler premise, it is also important to recognize the context in which it is being conducted. As the blog post describing the research concludes:

We are focused on using robotics work that will not only lead to more capable robots but will also push the limits of AI over the years and decades to come. If we want to move closer to machines that can think, plan, and reason the way people do, then we need to build AI systems that can learn for themselves in a multitude of scenarios — beyond the digital world.

As Facebook continually works on expanding its influence from its walled garden of apps and services into the rich but unstructured world of your living room, kitchen, and office, its AI agents require more and more sophistication. Sure, you won’t see a “Facebook robot” any time soon… unless you count the one they already sell, or the one in your pocket right now.

This clever transforming robot flies and rolls on its rotating arms

There’s great potential in using both drones and ground-based robots for situations like disaster response, but generally these platforms either fly or creep along the ground. Not the “Flying STAR,” which does both quite well, and through a mechanism so clever and simple you’ll wish you’d thought of it.

Conceived of by researchers at Ben-Gurion University in Israel, the “flying sprawl-tuned autonomous robot” is based on the elementary observation that both rotors and wheels spin. So why shouldn’t a vehicle have both?

Well, there are lots of good reasons why it’s difficult to create such a hybrid, but the team, led by David Zarrouk, overcame them with the help of today’s high-powered, lightweight drone components. The result is a robot that can easily fly when it needs to, then land softly and, by tilting the rotor arms downwards, direct that same motive force into four wheels.

Of course you could have a drone that simply has a couple wheels on the bottom that let it roll along. But this improves on that idea in several ways. In the first place, it’s mechanically more efficient since the same motor drives the rotors and wheels at the same time — though when rolling the RPMs are of course considerably lower. But the rotating arms also give the robot a flexible stance, large wheelbase, and high clearance that make it much more capable on rough terrain.

You can watch FSTAR fly, roll, transform, flatten, and so on in the following video, prepared for presentation at the IEEE International Convention on Robotics and Automation in Montreal:

The ability to roll along at up to 8 feet per second using comparatively little energy, while also being able to leap over obstacles, scale stairs, or simply ascend and fly to a new location give FSTAR considerable adaptability.

“We plan to develop larger and smaller versions to expand this family of sprawling robots for different applications, as well as algorithms that will help exploit speed and cost of transport for these flying/driving robots,” said Zarrouk in a press release.

Obviously at present this is a mere prototype, and will need further work to bring it to a state where it could be useful for rescue teams, commercial operations, and the military.

Making industrial chemicals “green” requires a lot of renewable electricity

A factory that makes industrial chemicals from fossil fuels.

When we think about climate change, we most often think about emissions from two sectors: energy and transportation. But industry makes a big contribution to climate change, too. Industrial emissions come from a lot of different things, including the manufacture of common chemicals. Often, these chemicals are made by reforming fossil fuels using heat that’s also provided by burning fossil fuels.

Overall, the chemical industry consumes about 10 percent of global final energy, according to the International Energy Agency.

In a recent PNAS paper, researchers from universities in Germany and California tried to estimate how effectively the chemical industry could decarbonize and whether such a decarbonization is likely.

Read 11 remaining paragraphs | Comments

Optical neural network at 50zJ per op? Nope, but it’s still a good idea

Optical neural network at 50zJ per op? Nope, but it’s still a good idea

Artificial intelligence (AI) has experienced a revival of pretty large proportions in the last decade. We’ve gone from AI being mostly useless to letting it ruin our lives in obscure and opaque ways. We’ve even given AI the task of crashing our cars for us.

AI experts will tell us that we just need bigger neural networks and the cars will probably stop crashing. You can get there by adding more graphics cards to an AI, but the power consumption becomes excessive. The ideal solution would be a neural network that can process and shovel data around at near-zero energy cost, which may be where we are headed with optical neural networks.

To give you an idea of the scale of energy we’re talking about here, a good GPU uses 20 picoJoules (1pJ is 10-12J ) for each multiply and accumulate operation. A purpose-built integrated circuit can reduce that to about 1pJ. But if a team of researchers is correct, an optical neural network might reduce that number to an incredible 50 zeptoJoules (1zJ is 10-21J).

Read 15 remaining paragraphs | Comments

NASA’s full Artemis plan revealed: 37 launches and a lunar outpost

A Blue Moon lander, built by Blue Origin, with an ascent vehicle (built by another company) on top.

In the nearly two months since Vice President Mike Pence directed NASA to return to the Moon by 2024, space agency engineers have been working to put together a plan that leverages existing technology, large projects nearing completion, and commercial rockets to bring this about.

Last week, an updated plan that demonstrated a human landing in 2024, annual sorties to the lunar surface thereafter, and the beginning of a Moon base by 2028, began circulating within the agency. This graphic, shown below, provides information about each of the major launches needed to construct a small Lunar Gateway, stage elements of a lunar lander there, fly crews to the Moon and back, and conduct refueling missions.

This decade-long plan, which entails 37 launches of private and NASA rockets, as well as a mix of robotic and human landers, culminates with a “Lunar Surface Asset Deployment” in 2028, likely the beginning of a surface outpost for long-duration crew stays. Developed by the agency’s senior human spaceflight manager, Bill Gerstenmaier, this plan is everything Pence asked for—an urgent human return, a Moon base, a mix of existing and new contractors.

Read 15 remaining paragraphs | Comments

Wandering Earth: Rocket scientist explains how we could move our planet

Wandering Earth: Rocket scientist explains how we could move our planet

In the Chinese science fiction film The Wandering Earth, recently released on Netflix, humanity attempts to change the Earth’s orbit using enormous thrusters in order to escape the expanding Sun—and prevent a collision with Jupiter.

The scenario may one day come true. In five billion years, the Sun will run out of fuel and expand, most likely engulfing the Earth. A more immediate threat is a global warming apocalypse. Moving the Earth to a wider orbit could be a solution—and it is possible in theory.

But how could we go about it and what are the engineering challenges? For the sake of argument, let us assume that we aim to move the Earth from its current orbit to an orbit 50% further from the Sun, similar to Mars’.

Read 17 remaining paragraphs | Comments