Today marks the start of the IEEE International Conference on Robotics and Automation (ICRA 2009) in Kobe, Japan, where researchers from around the world will gather to discuss the latest advances in robotics–from cutting-edge climbing machines to robots that politely ask for directions.
Researchers from the University of Pennsylvania will present the latest version of RiSE, a four-legged robot that can both scamper along the ground and rapidly climb a tree or a pole. RiSE V3 was designed and built at Boston Dynamics–the company behind the four-legged military robot BigDog. It has four legs, and tiny claws made from surgical needles that can dig into a vertical surface. The robot’s front legs are long enough to hug a telephone pole, and it can climb at 21 centimeters per second.
“RiSE V3 is the first general-purpose legged machine to achieve this vertical climbing speed,” says Daniel Koditschek, a professor of electrical and systems engineering at the University of Pennsylvania, who led the work. Because the robot can walk, climb, and rest quietly on a pole while conserving energy (watch a video), Koditschek says that it could “play an invaluable role in search and rescue, reconnaissance, surveillance, or inspection applications.”
Another mobile robot set to debut at the event is Adelopod, developed by researchers at the University of Minnesota. Adelopod, which is about the size of a video controller, doesn’t use legs or even wheels to get around. Instead, it flips itself over and over using a pair of 12-centimeter arms (video of Adelpod in action). This tumbling mode of locomotion is simple, saves energy, and doesn’t require complex hardware, say the researchers involved. “Given its size, it can go places that other robots cannot,” says Nikos Papanikolopoulos, director of the university’s Center for Distributed Robotics. The group has also developed the larger Loper robot, which can carry several Adelopods and scatter them throughout an area.
See robotic cars navigate around a small city.
See a car-based camera system identify pedestrians.
Watch researchers test the impact of collisions between robots and dummies.
Watch a robot make ice sculptures.
Watch a robot climb a telephone pole.
Researchers at the Institute of Automatic Control Engineering at the Technical University of Munich (TUM), in Germany, have designed a robot that can find its way around a city without GPS or preloaded maps. It does so by asking pedestrians for directions and using gesture tracking and voice recognition to interpret commands. It also uses human tracking, obstacle detection, and map building to guide itself around a busy city. “The novelty about our research is that we have a robotic system that uses human instructions as global waypoints for navigation in an outdoor environment,” says Andrea Bauer, one of the researchers at TUM. “The robot can retrieve missing route knowledge just like a person, by asking passersby.” Watch a video of the robot on TUM’s website here.
Another project that involves robots navigating around busy city streets will be demoed by researchers from Boston University. They have created a miniature city–complete with robotic cars–to test different approaches to control and navigation. The researchers’ Robotic Urban-Like Environment (RULE) system lets the cars understand a simple, high-level command by a human, such as “Take me to the grocery store”; demonstrations show that the robotic cars can not only reach their destination safely, but can also move into the correct lane, stop at red lights, and even park on their own (watch a video). Automated vehicle systems, such as the ones being built in Masdar and Heathrow, currently require some kind of track or magnetic guiding strip for navigation. “We wanted to give the robots the freedom to make choices by themselves as long as they are safe and accomplish whatever the human operator specified as a task,” says Calin Belta, a professor and lead researcher of the work.
Robotic cars will of course need to be able to spot, and respond to, unexpected dangers, and another system that will be presented at ICRA 2009 is designed to do this. It was developed by researchers at ETH Zurich, and others and can quickly identify pedestrians and other obstacles and predict their paths in order to avoid them. When mounted atop a car, the system can rapidly outline pedestrians even in areas full of traffic and clutter (watch a video). “The idea is to equip cars with vision systems that can oversee the traffic situation around the car and that can give an early warning for dangerous situations,” says Luc Van Gool, a professor at the Computer Vision Laboratory at ETH Zurich, who developed the system with his colleague Andreas Ess.
Making robots safer will be important if they are to find use in everyday life. Researcher at the German Aerospace Center (DLR) will describe experiments involving crash-test dummies designed to explore robot-human accidents. The researchers designed a robot that pulls itself back when it starts to detect an impact to the dummy’s chest or head (see a video of robot crash testing). “In order to provide really suitable methodologies for making robots safe, we need to understand what the relevant threats are by physical means,” says Sami Haddadin, a research engineer at DLR. “We aim to establish a testing protocol for robots which qualify them for use in human proximity. This would bring us to a point at which everyday human-robot collaboration becomes reality.”
Of course, no robotics conference would be complete without a few oddball robots and machines, and ICRA is no exception. Take, for example, a robot that mimics how a human spins pizza and another that picks up empty coffee cups from around an office. Researchers at McGill University, in Montreal, Canada, have even made a robot that can create ice sculptures on its own (watch a video). To do this, they modified a device called the Cobra, typically used for repetitive tasks like picking up objects from a conveyer belt, to lay down deposits of ice. The researchers say that this technique could eventually be used for rapid prototyping of other materials as well.
DeepMind’s cofounder: Generative AI is just a phase. What’s next is interactive AI.
“This is a profound moment in the history of technology,” says Mustafa Suleyman.
What to know about this autumn’s covid vaccines
New variants will pose a challenge, but early signs suggest the shots will still boost antibody responses.
Human-plus-AI solutions mitigate security threats
With the right human oversight, emerging technologies like artificial intelligence can help keep business and customer data secure
Next slide, please: A brief history of the corporate presentation
From million-dollar slide shows to Steve Jobs’s introduction of the iPhone, a bit of show business never hurt plain old business.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.