Every few weeks, it seems, a new story hits the headlines about the latest breakthrough in the field of robotics. Innovations are happening all the time, of course. The term "robotics" encompasses a global system of technology, industry and research. But certain kinds of robot stories break through into popular culture with alarming regularity.
Taken together, these stories can provoke a certain unease. Read enough of them in a row, and the question presents itself: Are we watching the robot revolution take place before our very eyes? Will our future robot overlords look back, 100 years from now, and regard these moments with digital fondness as they populate their own history books? (Or databases or holo-archives or whatever they have in store?)
Forthwith, a look at recent developments in robotics that could well prove historical, albeit for all the wrong reasons.
History has taught us that the key to a successful uprising is organization. Enter the RoboEarth project, billed as "a World Wide Web for robots." By way of a cloud-based network and database, RoboEarth allows robots to share information over the Internet. Data stored in the RoboEarth knowledge base might include maps for navigation, specific task instructions or images for object recognition.
Robots can also offload any resource-heavy computation from onboard systems to the cloud. RoboEarth program manager Markus Waibel, of the Swiss Federal Institute of Technology in Zurich, said the RoboEarth infrastructure helps robots learn from one another. "In our demonstrations, we showed that a humanoid robot could greatly speed up its task performance by using prior information ... stored in the RoboEarth database," he said.
"What's more, we could show that a second robot with different hardware and software could also execute this complex task faster if it had access to RoboEarth," Waibel said. "RoboEarth could allow robots to carry out useful tasks that were not explicitly planned for at design time."
Of course, when plotting a planetary revolution, it never hurts to be polite. Sara Kiesler, along with her colleagues at the Human-Computer Interaction Institute at Carnegie Mellon University, are studying how to make robots more personable.
Kiesler's team recently presented an interesting study at the International Conference on Human-Robot Interaction, held in Tokyo last month. By studying how expert chefs shared advice with students, they're hoping to program robotic dialogue systems in which robots can give instructions and advice to humans … politely.
"We wondered, if you're trying to be polite, how do you actually do this if you're a robot," Kiesler said.
It turns out that we humans unconsciously use certain rhetorical "hedges" in conversation -- phrases like "so" and "well" and "maybe we can try" -- to soften interactions when instructions are being delivered. Kiesler's research suggests that when robots incorporate such conversational tweaks into dialogue, they appear friendlier and less threatening to humans.
Dancers perform at at a salsa show in Cali, Colombia. Are the robots watching ... learning?Mark Harris/Getty Images
A crucial element of strategy in any insurrection is anticipating the enemy's next move. A recent project at Boston University's Intelligent Mechatronics Lab (IML) has introduced to the world what is surely a brand new kind of automaton –- the salsa bot.
Researchers began by mapping the movements of actual salsa dancers, then programming dance-partner robots with four basic beginner moves. The robots use motion sensors to identify and anticipate the moves of the dance in real-time.
The idea is to build machines that can respond independently in rapidly changing or dangerous situations such as rescue missions, said IML director John Baillieul. "The ultimate goal is to understand human reaction to gestures, and how machines may react to gestures."
YouTube screen capture
As the Bond Babes in 007 movies have taught us, there are softer and more devious ways to get close to your enemy. For several years now, the decidedly strange Project Aiko has been popping up in news stories about odd developments in the robotics world. Developed by Canadian inventor Le Trung, Aiko is a humanoid companion robot that can read newspapers, check the weather, discern among individual faces and ... um ... be tickled.
"It actually starts back in the '70s during my childhood, where I spent a lot of time watching Japanese anime, which often featured robot themes," Trung writes on his official Project Aiko homepage. "It is currently impossible to build robots that will have the same feelings and emotions as us, but we can start by building a robot that looks human and does artistic human things such as read music and sing for us."
Project Aiko has never gotten much traction in the scientific world -- demos take place at hobby shows, not international conferences. But the story has gotten a lot of play in the media, and clearly reflects one man's vision of our robotic future.
The annals of science fiction are, of course, packed with robot apocalypse scenarios. As a concept, it's been provoking cultural anxiety and generally giving us heebie-jeebies for decades. Countless films and novels posit a future that seems more familiar every day: A global network of computers (!) supplant human society by way of lifelike androids (!!) and lethal robotic drones (!!!).
Ah, well. If the Earth is overrun, we can always retreat to our evolutionary womb in the oceans, right?
Not so fast. According to reports from just this month, Virginia Tech College of Engineering researchers have developed an autonomous robotic jellyfish nearly six feet long and weighing in at 170 pounds. The prototype, named Cyro, is part of a nationwide project funded by the Office of Naval Research and the U.S. Naval Undersea Warfare Center.