Welcome
To
Mechatronics World

Sunday, 20 May 2018

DARPA and Drone Cars: How the US Military Spawned Self-Driving Car Revolution



DARPA and Drone Cars: How the US Military Spawned Self-Driving Car Revolution
In the 2004 DARPA Grand Challenge, none of the robotic, self-driving cars successfully finished the challenging course from Barstow, Calif., to Primm, Nev.
Credit: DARPA

The Defense Advanced Research Projects Agency (DARPA), the arm of the U.S. Department of Defense responsible for advancing military technology, is celebrating the 10th anniversary of the DARPA Grand Challenge, an ambitious, first-of-its-kind race between robotic, self-driving cars.
The Grand Challenge, which occurred on March 13, 2004, involved 15 self-driving carstrying to navigate a 142-mile (228 kilometers) course between Barstow, Calif., and Primm, Nev. The contest, while seemingly primitive at the time, has since led to vast improvements in robotic technology, and has demonstrated the value of competitions as a way to drive innovation and collaboration among scientists and engineers.
The DARPA Grand Challenge was designed to foster and accelerate the development of autonomous vehicles that could one day be used to transport cargo and other military supplies into combat zones without endangering the lives of human drivers, according agency officials. The contest was also DARPA's first major attempt to use prize money as an incentive for innovation within the research community. [Photos: The Robotic Evolution of Self-Driving Cars]
Legacy of the Grand Challenge
As part of the DARPA Grand Challenge, the first team to pass a series of qualification tests and complete the course in less than 10 hours would win the $1 million top prize.
But, the contest revealed an industry in its infancy, because none of the cars managed to finish the rugged desert course. In fact, the highest-scoring vehicle traveled only 7.5 miles (12 km), and none of the teams claimed the $1 million prize.
Yet, the DARPA Grand Challenge proved to be an important milestone.
"That first competition created a community of innovators, engineers, students, programmers, off-road racers, backyard mechanics, inventors and dreamers who came together to make history by trying to solve a tough technical problems," Lt. Col. Scott Wadle, DARPA's liaison to the U.S. Marine Corps, said in a statement. "The fresh thinking they brought was the spark that has triggered major advances in the development of autonomous robotic ground vehicle technology in the years since," Wadle said.
DARPA's second Grand Challenge was held in the fall of 2005, with 195 teams vying for a $2 million prize. This time, five vehicles successfully completed the 132-mile (212 km) course across southern Nevada.
A team from Stanford University claimed the top prize, after their car "Stanley" finished the course with a time of 6 hours and 53 minutes, DARPA officials said.
In 2007, DARPA kicked off its third competition, the Urban Challenge. The contest involved driverless cars trying to maneuver through a complex course in Victorville, Calif., demonstrating the ability to manage obstacles and other moving traffic while simultaneously obeying the state's traffic regulations, DARPA officials said.
Six of the 11 teams completed the course, and a team from Carnegie Mellon University in Pittsburgh won the $2 million prize.
Foundation for the future
The challenges have helped foster new and useful technologies for the military, including an autonomous utility vehicle used by the Marine Corps that was developed by one of the six finishers of the Urban Challenge, according to DARPA officials.
The contests have also provided a model for other similar prize-based competitions, including the Spectrum Challenge, which involves teams competing to demonstrate how best to use reliable radio communication amid interfering signals.

Saturday, 19 May 2018

Tiny Robot Flies Like a Jellyfish




Tiny Robot Flies Like a Jellyfish
A new robot that measures only 3 inches (8 centimeters) long flies through the air like a jellyfish swims.
Credit: Leif Ristroph

A new teeny-tiny robot flies through the air like a jellyfish swims.
The jellyfish flier is a strange sight — it looks a little bit like a Chinese lantern that's developed a hankering for the open skies — but its unique design keeps it from tipping over without the use of sensors or external controls. That talent could make it handy for maneuvering in small spaces, said its inventor Leif Ristroph, a postdoctoral researcher at New York University.
"What's cool is you can actually build these flying things yourself," Ristroph told LiveScience. "All the components I used to make this, they cost about $15 and they're available on hobby airplane websites." [See Video of the Flying Jellyfish Robot]
Many small robots take insects as their inspiration, such as the dragonfly-like DelFly, a Dutch drone that weights only 0.7 ounces (20 grams). But insect flight is notoriously unstable, Ristroph said. Instability is no big deal for bugs — they have nervous systems that can correct their wing motions when they begin to tip. Insect-inspired robots need sophisticated sensor systems to do the same.
"The real number-one issue in flying is stability," Ristroph said. "How do I keep upright the whole time?"
Ristroph wanted to know if he could invent a winged flying machine that was inherently stable, no sensors or artificial nervous system needed. He tinkered with five or 10 designs, he said, including one inspired by an umbrella. In principle, he said, opening and closing an umbrella rapidly should generate lift and send the object flying.
The flying jellyfish
He ended up creating a conelike machine made of four wings approximately 3 inches (8 centimeters) long, surrounding a small motor. The wings flap up and down not-quite-simultaneously. In practice, Ristroph said, the motion is very similar to what a jellyfish does to propel itself through water.
"It has a bell that contracts and squirts water downwards," Ristroph said. The jellyfish flier does the same with air.
The magic of the new robot is that it is automatically stable. If it begins to tip, the motion of the flapping wings naturally corrects its course, keeping it upright as it ascends, hovers and moves horizontally.
The prototype does not include a battery, so the robot still requires a wire for power. More engineering work will be necessary to get power and a radio receiver onboard so that an operator can control the flying bot from afar, Ristroph said.
"Part of what we're trying to do is get people excited about this design so that we can work with them and improve upon it," he said. He and his colleagues detail their invention today (Jan. 14) in the Journal of the Royal Society Interface. They also presented the robot at a fluid dynamics meeting in November.

Real-Life 'RoboCop' May Be Coming to a Street Near You



Real-Life 'RoboCop' May Be Coming to a Street Near You
A bullet-shaped robot that stands about 5 feet (1.5 meters) tall, the K5 "security guard" looks a lot like the droid R2-D2 from the "Star Wars" films. And it may be able to help stop crimes before they happen.
Credit: Knightscope

Robot security guards are staples of most futuristic sci-fi movies, video games and TV shows. They exist in real life as well, though the sight of a security robot patrolling the streets is far from common.
The K5 Beta, a just-unveiled prototype from California-based company Knightscope, might change all that.
A bullet-shaped robot that stands about 5 feet (1.5 meters) tall, the K5 looks a lot like the droid R2-D2 from the "Star Wars" films. The K5 might not have all the features of its counterpart from a galaxy far, far away. But what it does have, Knightscope representatives said, is an onboard sensor array that can see, hear, touch and smell its surroundings. [5 Reasons to Fear Robots]

The K5 can also combine that sensory data with "existing raw business, government and crowdsourced social data sets, and subsequently assigns an alert level that determines when a business, community or authorities should be notified of a concern," according to Knightscope's press release.
In other words, the K5 is supposed to combine its observations with public data on the social and financial statistics of its surroundings, and use the information to predict if, when and where a crime is likely to occur.
If the K5 does detect that an "incident" is occurring, it makes all of its sensor data publicly available via Wi-Fi, "to allow the entire community to review the information transparently and contribute additional relevant, real-time information," Knightscope representatives said.
Crowdsourced crime investigations have their pros and cons, as anyone who followed the immediate aftermath of the Boston Marathon bombingknows well. Using Twitter, Boston residents were able to inform others about the events surrounding the attack and help identify the alleged bomber, but not before an innocent person was publicly named as a suspect. Presumably, the K5's software will be able to sift through crowdsourced data to find the most valuable information.
The company was founded in response to another deadly incident that gained national attention: the shooting massacre at Sandy Hook Elementary School in Newtown, Conn., in December 2012. 
"Our long-term goal is to show that, with a combination of hardware, software and community involvement, we will, together, be able to cut crime," Knightscope CEO William Santana Li said in a statement.
Many people aren't comfortable with the idea of human police officers carrying video cameras or amassing a private database of footage, so it's unlikely that the public would approve of camera-toting robots. The K5's ability to make its collected data publicly available is intended, in part, to assuage those concerns, as Li told The New York Times. The idea is that people will feel more comfortable if the data the K5 collects is in everyone's hands, and not just the police or a private group.
What's more, Knightscope representatives said the K5 will be able to save money. The robot operates at a cost of about $6.25 per hour, the New York Times reports — more than a dollar below the U.S. minimum hourly wage.

Featured post

Reading Robot Minds with Virtual Reality

Figuring out what other people are thinking is tough, but figuring out what a robot is thinking can be downright impossible. With no bra...

Popular Posts