More than 50 leading AI and robotics researchers have joined a boycott of South Korea’s KAIST university over the institute’s plans to help develop AI-powered weapons. The boycott was announced ahead of a UN meeting set in Geneva next week to discuss international restrictions on so-called “killer robots.” It marks an escalation in tactics from the part of the scientific community actively fighting for stronger controls on AI-controlled weaponry.
The boycott was organized by Professor Toby Walsh of the University of New South Wales, who warned in a press statement that the race to build autonomous weapons had already begun. “We can see prototypes of autonomous weapons under development today by many nations including the US, China,…
When a person screws up we call it human nature. So what does it mean when a machine that’s trying to imitate our intelligence makes a mistake? According to the doomsayers, it means robots could attack us because of faulty reasoning – and that’s scary. But, it’s hard to fear a machine that can be defeated with tropical fruit. That’s why we’ve gathered some of the best robot fails we could find to remind everyone we’re still in charge. For starters, who could forget the Boston Dynamics’ Atlas robot demonstration? The company showed off its new technology at The Congress of…
As more and more industrial robots come online, there's an ever greater risk of hackers gaining access to them and taking control of their abilities. In some cases that could mean damage to a product or manufacturing flow, but in other cases, it coul… Engadget RSS Feed
Giant robots duke it out in the middle of major cities. Humans inside robotic exoskeletons control them just by thinking about it. The robots hack each other with massive saws or fling stacks of cars at one another.
Yes, Pacific Rim: Uprisingis the most popular movie in the U.S. right now. In it, humans must pilot enormous robots, called Jaegers, to ward off other, evil enormous robots.
The bots themselves seem pretty high-tech. But there’s an element that may not be as apparent: our real-life technology is actually much more advanced.
That’s according to Robin Murphy, a professor of computer science and engineering at Texas A&M University, in an article published today in the journal Science Robotics.
If robots like the Jaegers seem like a staple of sci-fi everywhere, that’s because they pretty much are, Murphy writes. In the 1960s and 70s, the U.S. military, in partnership with GE, made the first real-life exoskeleton, and dozens more have come out in years since.
Through it all, engineers have learned a few things — things that have fallen by the wayside in Pacific Rim. Here’s a brief breakdown of what Murphy saw to be lacking in the Jaegers, and what scientists have already learned about how to do it better.
Size
The Jaegers of Pacific Rim clock in at an impressive height of about 76 meters (250 feet). However, controlling such massively complex robots in reality simply “cannot be done,” according to Murphy. Today’s researchers are far more likely to focus on smaller robotic exosuits, similar in size to those worn by Tony Stark in Iron Man or Tom Cruise in Edge of Tomorrow.
Walking
It looks really hard to get a Jaeger to walk or run. That’s because, in the movies, the massive bots mirror every step their pilot takes — a process that is much more complicated than it has to be.
“In reality, locomotion is becoming one of the easiest functions to totally delegate to a robot,” writes Murphy.
Think about the Boston Dynamics robots. Engineers simply communicate a speed and direction to one, and it handles the process of lifting and lowering each foot.
Use
In Pacific Rim, pilots use Jaegers as weapons against hostile aliens. In reality, we’re far more likely to design robotic exoskeletons to allow humans to do everyday tasks more easily and safely. Think the power loader used in the movie Aliens to lift heavy materials. You know, before Ellen Ripley uses it as a weapon against a hostile alien.
In fact, Murphy notes, Hollywood has largely ignored one of the most likely uses for human-powered robots: healthcare. Researchers have used exoskeletons to help people with spinal cord injuries walk again.
Our current technology might not make you say “dude that’s awesome!” as much as the stuff in the movies. But as Murphy notes in her article, our current tech actually is awesome.
We were warned, time and time again, by one expert after another. AI doesn’t play fair, it can’t make moral decisions, and it relies entirely on humans for ethics. When those humans are (allegedly) Alex Nix, Steve Bannon, and Dr. Aleksandr Kogan we’re all in a lot of trouble. Cambridge Analytica (allegedly) took bad science, stolen data, and machine learning and used it to create a toxic stew designed specifically to radicalize the American people. The company’s success is no more admirable than a serial killer’s bodycount. And no matter Facebook’s level of involvement, it’s now tainted. Even if it turns out…
Makeup-users, get ready. AI is ready to tell us we look ugly.
The app ModiFace scrutinizes your skin type, points out what’s wrong with your face, (you’ve got wrinkles here, a spot there), and suggests cosmetics that might be able to repair or cover those flaws.
If you just thought, “Hey, making people feel insecure is a great way to get people to buy cosmetics,” you’re not alone. French cosmetic giant L’Oreal recently purchased the Canadian company for an undisclosed (but likely astronomical) sum.
In theory, the tool sounds kind of cool. The company has published more than 200 scientific studies and holds 30 patents in development of this technology, so there’s reason to believe it works better than some of the other makeup apps already out there (ModiFace would probably become the best-known of these once L’Oreal makes it ubiquitous). And because it’s an app, it doesn’t require investment in a whole new piece of hardware, the way Amazon smartmirror would.
Plus it sounds pretty convenient. Trying a new foundation without having to go to the store and covering my hand with poorly-colored product? Yes, please.
But there’s an obvious downside to ModiFace: it provides users with a limited definition of what it means to look beautiful.
That standard is getting narrower all the time thanks to social media. Despite some effort to expand beauty standards to people of different colors, shapes, and sizes, the popularity of sites like Instagram and Snapchat make the same old beauty standards seem attainable and immediate, the airbrushing and photoshopping seem less obvious.
“We are on the verge of a very, very serious problem,”Jane Cunningham, the author of Britishbeautyblogger.com, told The Guardian in 2015. “The world of vloggers and YouTubers has created a perverse, homogenized sense of distorted ‘beauty’ with no diversity or reality… Society is losing all perspective on the diversity of beauty and it’s contributing to an alarming growth in dysmorphia.” Recent scientific literaturehas confirmed this assessment.
Consumers, especially women, don’t need another faceless entity pointing out every little flaw. That’s especially true for women of color, for whom facial recognition software can be inaccurate. Might ModiFace direct non-white users towards more limited products for their skin tones, or push them to ascribe to a particular aesthetic they may not agree with? Well, it wouldn’t be the first time. And it’s not clear from ModiFace’s marketing materials whether the tool works for non-cis women — a growing market of makeup consumers (in 2016 L’Oreal itself used a male model to advertise cosmetics).
A tool like ModiFace might be useful for some makeup users, or convenient even. But for others, it will just be another entity telling them they’re not young enough, not white enough, not pretty enough.
The past year was marked with increasing concerns about the effects artificial intelligence and automation will have on human employment and labor. Not a month goes by without one (or ten) articles warning about robots driving humans into unemployment and causing economic and social chaos. But maybe we’re looking at things from the wrong perspective, through the lens of the laws that govern our current world. In a world where robots perform all tasks, many of the rules that we live by today might become obsolete. And it might not be such a bad thing. The value of human labor In today’s…
Parcel delivery and logistics firm DHL said today that it is piloting a range of new connected technologies in the US, including collaborative robots, AI, and augmented reality glasses.
The move comes in the wake of retail and Web services hyper-company Amazon’s recent announcement that it is moving into deliveries in America.
“As consumer expectations are rapidly evolving due to a number of major trends, including e-commerce, urbanisation, and sustainability in particular, logistics providers are being challenged to provide more flexible and efficient services,” said DHL today.
DHL Supply Chain has introduced robotics in its warehouse operations – including LocusBots and Sawyer collaborative robots – along with augmented reality glasses. The mix of technologies is designed to improve ‘smart picking’, productivity, and order fulfilment.
LocusBots work autonomously in warehouses to find and move goods, as this video explains.
DHL Supply Chain President of Retail, Jim Gehr said: “Innovative technology is becoming an extremely important element in e-commerce logistics because of changing customer demands.
We are using software technology to speed order flow and to more efficiently organise distribution centres. Hardware solutions like LocusBots allow for faster picking. What we look for is a smart combination of software and hardware.
Last year, the Deutsche Post DHL group trialled autonomous robot deliveries in Germany using its PostBOT vehicles (pictured above), and has also been trialling ‘Parcelcopter’ drones.
Meanwhile, the company’s DHL Express division has been piloting artificial intelligence in its customer applications to further enhance its responsiveness and customer service.
The company said today that it has “registered a 10 percent increase in shipment processing accuracy through increased automation at its hubs and gateways”, while its on-demand delivery online service has increased first-time delivery success for e-commerce shipments from 80 percent to 92 percent.
Last-mile services
As part of the delivery giant’s widespread internal transformation via IoT and digital technologies, its DHL eCommerce unit has also launched ‘DHL Parcel Metro’, a new service that helps online retailers satisfy customer demands for same- or next-day delivery.
Parcel Metro uses customised software that allows DHL eCommerce to create a ‘virtual delivery network’ of local or regional contract couriers along with crowd-sourced providers. This ensures maximum flexibility and capacity over the last mile, it said today.
The new service is now available in Chicago, New York, and Los Angeles, and will be launched in Dallas, Atlanta, San Francisco, and Washington DC later in the year.
Bobby Shome, Global Business Development Director at delivery management specialist Centiro, explained the thinking behind the strategy: “DHL’s launch of a same-day delivery service for US online retailers drives home just how important the last mile has become in retaining and attracting customers.
“These new delivery options from DHL will allow smaller online businesses in the US to successfully compete with larger, more established retailers, like Amazon, which recently launched its own delivery service.
“For all the progress that’s being made, however, retailers mustn’t lose sight of the fact that they risk damaging relationships with customers if the delivery service does not meet expectations. As such, it is critical for retailers to retain visibility over the delivery process, and ensure they have the technology in place to make this possible, regardless of the promises being made by carriers.”
Internet of Business says
The digital and connected transformation of retail, the supply chain, and deliveries continues apace, and is centred on customer need. As more and more people opt for the convenience of shopping with a click, any old-fashioned lags in the system rapidly become unacceptable. However, all organisations should be wary of any automated system that increases the end-customer’s workload, even as it speeds up the delivery experience.
Farmers in the Australian state of Victoria have been promised a $ 15 million financial and technological boost that is expected to be “a game-changer” for the agricultural sector.
The funding is intended to help farmers deploy robotics, wireless networks, sensors, and analytics solutions to capitalise on the enormous potential of connected technologies to make agriculture smarter and more efficient.
With rising labour costs, and with utilities and supplies bills soaring, farmers are increasingly looking at how emerging technologies can help sustain their businesses. However, many come with heavy upfront costs: networks need to be installed and expensive hardware needs to be purchased.
From harvesting robots to drone-assisted aerial surveys and field sensors, the vision of the connected farm is an appealing one. But getting to that stage can require a joint effort, often between one farm and another, but also between the state and agricultural communities.
The Victoria government has announced $ 15 million of IoT-focused funding, and trials will begin in the regions of Maffra, Birchip, Serpentine, and Tatura in July. The state previously made a $ 12 million investment into IoT and agtech as part of a demonstration trial in 2016-17.
Connecting Victoria’s farms
According to a statement from the Victoria government, the funding will be put towards a range of digital innovations. These include robotics, and the development of IoT networks, wireless technology, biotechnology, and virtual fencing.
Sensors and IoT networks will be installed for both farmer and public access. These will provide insights on weather and soil conditions, creating benefits across the dairy, grain, sheep, and horticulture sectors.
Farms right across the state will install sensors and connected devices as part of the project. The data generated will be uploaded to a central system for analysis, and be accessible via a software platform that will enable farmers throughout the region to make more informed decisions.
Victoria’s minister for agriculture, Jaala Pulford, said: “Victoria is the agtech hub of Australia, and with this investment, we are looking at being a world leader. Our farmers deserve the very best tools to get the job done and digital innovation is at the heart of this. We’re proud to play our part in making this a reality.”
Jaala Pulford, Victoria’s agriculture minister
“This Internet of Things Demonstration Trial is an important step in maximising technology to help provide Victorian agriculture with a competitive advantage on a global scale,” she added.
Despite being one of the oldest industries, farming is at the forefront of IoT, robotics, and big data applications. The reasons are manifold, and include the challenges of seasonal labour, soaring costs, climate change, unpredictable weather, changing consumer demands, and international competition.
The IoT’s mix of smart hardware, AI, sensors, and data analytics mean that farmers can gain real insights into how efficiently and sustainably their farms are operating, gather data over time – and in real time – and ensure that crops and livestock are developing in ideal conditions and are being properly fed and watered.
For example, drones can offer multispectral imaging from the air, and direct autonomous farm machinery to areas that need irrigation and fertilisation. Meanwhile, sensor networks in the ground can monitor crop and climate conditions and, via AI, machine learning, and analytics, help farmers to build up a comprehensive and predictive picture of how well the land is performing.
Meanwhile, the same technologies are helping farms move into cities, closer to the mouths that need feeding. High-tech startups such as Aerofarms are using sensors and big data technologies to grow crops indoors, using smart lighting and chemical engineering to create the ideal conditions for crops to flourish.
Astronauts need a lot of help from robots. Whether it’s the crew aboard the International Space Station (ISS) or future recruits on missions exploring the depths of the Solar System, robots help us to complete tasks beyond our human capabilities. But what type of robots are best suited for these cosmic jobs? One NASA engineer thinks that we should turn to Star Wars for inspiration.
In the popular saga, Astromechs are repair droids that act as autonomous mechanics aboard ships. They are also capable of fighting, piloting spacecraft, and just about everything else. You might recognize astromechs like BB-8 or R2-D2. At NASA, these types of robots are called caretaker robots. In Science Robotics, NASA engineer W. Kris Verdeyen recently argued that the ideal robot for missions in space would have the capabilities of an astromech (or caretaker robot), but with a less problematic body.
Even on a film set, Verdeyen observed, rolling droids like BB-8 are not ideal for difficult, desert terrain. NASA has tested out humanoid robots like Robonaut and Robonaut 2, which flew to the ISS, but so far they have not been able to compete with the enormous variety of functions of astromechs. Still, because of its humanoid frame, “Robonaut, either the original or Robonaut 2, has demonstrated use of drills, torque wrenches, surgical equipment, air-quality testers—an impressive array of hand tools that may be necessary for an on-orbit repair,” Verdeyen wrote. He suggests that the ideal robot for such missions would have the capabilities of an astromech, or caretaker robot, but with a humanoid frame.
One of the primary functions that make astromechs so useful is their “creativity” — something that has been sorely lacking in NASA’s humanoid robots. They can autonomously solve problems that they haven’t been preprogrammed to handle. When in space, things never go exactly as planned, and an inability to respond quickly will threaten the safety of a mission. The ideal robot would be able to adapt to unexpected situations.
Verdeyen says that, to try to combine a physically capable robot with these qualities, NASA has been exploring what it calls “embedded intelligence.” They want to take robots like Robonaut 2 and equip them with artificial intelligence (AI) and a hefty database of basic knowledge to kickstart them.
It is unclear whether or not these ideal space robots would be as charismatic as their on-screen counterparts, but it’s possible they will be just as handy. “Star Wars got a lot of the functions of space robots right. NASA’s space robots will need to be able to repair spacecraft semiautonomously, just like R2D2 and BB-8,” Verdeyen said. “They will need to solve problems in and store knowledge about their spacecraft, just like R2D2 and BB-8. But, at least for the time being, they will look a lot more like C3PO and K-2SO.”
As more and more industrial robots come online, there's an ever greater risk of hackers gaining access to them and taking control of their abilities. In some cases that could mean damage to a product or manufacturing flow, but in other cases, it coul…
