OTHER than the People’s Liberation Army, Foxconn may well be China’s largest individual employer, and certainly its most important. The secretive electronics manufacturer, whose prestigious clients include Apple, has a workforce of more than 1m, including over 500,000 in one vast factory in Shenzhen.
Over the past decade Foxconn’s success has epitomised China’s ability to take elegant designs from high-wage countries and turn them out cheaply in huge quantities. Initially applauded for its ability to create vast numbers of jobs, the company’s success has recently come to be seen in a harsher light. Last year there was a spate of employees at the Shenzhen plant committing suicide; in the latest such case, a 21-year-old worker threw himself off a building [ http://www.channelnewsasia.com/stories/afp_asiapacific_business/view/1141979/1/.html ] in late July. In May an explosion at a new factory in Chengdu killed three more employees [ http://www.bbc.co.uk/news/business-13476800 ] and, it is believed, caused delays in production of Apple’s iPads.
To pacify its increasingly restive workers, Foxconn has repeatedly bumped up their wages, improved facilities, provided counselling and swathed its factories with nets to catch anyone leaping from a window. All this has resulted in higher costs, and signs that the company’s hitherto hugely successful business model has run its course. At a closed retreat in late July, Terry Gou, the chief executive of the company (which is also known as Hon Hai) unveiled a plan to replace a huge amount of human labour with robots by 2013.
In its public statement on the move, Foxconn talked about moving the more than 1m workers “higher up the value chain beyond basic manufacturing work” and of its “desire to move workers from more routine tasks to more value-added positions in manufacturing such as research and development, innovation and other areas that are equally important to the success of our operations.” But automation on the scale it is talking about would surely mean some of those human workers losing their jobs.
To a large extent, China’s recent economic development has been about reversing the rich-world trend towards automation—that is, using labour to replace capital in the manufacturing chain. Wages are now rising rapidly as a result of the demand for skilled labour (and because of big increases in the legal minimum wage), so one would expect this to make firms adjust their mix of labour and capital.
Many other, smaller, Chinese manufacturers have begun making similar moves, swapping increasingly costly labour with a bit more capital equipment. Foxconn is proposing to do it all (pardon the inevitable pun) in one great leap forward. But it may find this quite a challenge: its expertise has been in quietly running well-regimented armies of people making goods for highly visible companies. It is not known to have any particular skills in creating highly automated production lines; and moving in this direction will put it in direct competition with companies that do.
Still, it may have been an easy decision for Mr Gou. His highly image-conscious customers are bound to have been worried by the spate of horror stories about workers at Foxconn. If he did not change his production methods, the customers might feel obliged to look for another supplier.
Manufacturing experts and economists have been debating for some while now whether China’s rising labour costs and skill shortages might spell an end to the cheap “China price [ http://www.economist.com/node/18805862 ]”, leading global consumer-goods companies to shift elsewhere in Asia for their low-cost production—or even bring it back to their home countries. Certainly, some of the West’s strongest (and most highly automated) manufacturers, including Germany's Mittelstand firms [ http://www.economist.com/node/21524922 ], have proved surprisingly resilient to the pressure from cheap Chinese labour.
Others are sceptical: they say that Chinese firms are proving so adept at producing in ever greater quantities with fewer hands that they are set to remain fearsome competitors. Either way, the Chinese authorities will presumably be monitoring all of this very closely, and worrying about the prospect of the country’s industrial progress leading, as it did in the West, to large numbers of relatively well educated and articulate factory workers being thrown on to the streets.
Swarmanoid is a heterogeneous robot swarm in which different groups of robots have different capabilities: some robots are specialized in manipulating objects and climbing, some in moving on the ground and transporting objects, and some in flying and observing the environment from above. This video presents the Swarmanoid project, a 4 year research project coordinated by Marco Dorigo and funded by the Commission of the European Union.
This video won the Best Video Award at the AAAI-11 AI Video Competition.
You could call it Mission Impossible: Robot Library Heist. An army of flying, rolling, and climbing robots have been taught to work together to find and snatch a book from a high shelf.
In a striking display of military-like precision, the robotic team, dubbed the "Swarmanoid [ http://www.swarmanoid.org/ ]", attacks the problem with flying "eye-bots" and rolling "foot-bots". A "hand-bot" then fires a grappling hook-like device up to the ceiling and scales the bookshelf. Footage of the experiment, conducted by Marco Dorigo [ http://iridia.ulb.ac.be/~mdorigo/HomePageDorigo/ ] at the Free University of Brussels, Belgium, and colleagues, won the video competition at the Conference on Artificial Intelligence [ http://www.aaai.org/Conferences/AAAI/aaai11.php ] in San Francisco earlier this week.
It's a simple demonstration, but in the future Dorigo says the robots could be tasked with more difficult and important tasks. For instance, a bot team equipped like firefighters could wait in a building and spring to action when disaster strikes.
Six years ago, a swarm of identical rolling robots worked together to push a heavy object around a room [ http://www.newscientist.com/article/dn10319-robot-swarm-works-together-to-shift-heavy-objects.html ]. Integrating flying and climbing robots into the swarm allows it to explore its world faster, locate its objective, and perform more intricate tasks. Dorigo and his colleagues have so far amassed an army of 30 foot-bots, 10 eye-bots and eight hand-bots.
"It's not common to do a practical demonstration with heterogenous swarms," says Lynne Parker [ http://web.eecs.utk.edu/~parker/ ] of the University of Tennessee in Knoxville. "A lot of demonstrations are really basic lab demonstrations. They've really made a good effort to show how all [the robots] come together."
A swarm is born
When the Swarmanoid is first switched on, it knows nothing about its environment – a series of rooms with magnetic ceilings. One by one, flying eye-bots explore the rooms, searching for the target. Each one anchors itself to the magnetic ceiling once it's almost out of communication range from the other robots.
Once an eye-bot has spied the target, it signals its hanging comrades to activate the foot-bots. These rolling minions form a column from a central bot nest to the target, creating a communication chain for the hand-bot escort team.
Two foot-bots snap onto the hand-bot, escort it to the bookshelf, and release it. Then the hand-bot launches a magnet-tipped string to the ceiling and climbs the side of the bookshelf hand-over-hand, using the line for stability.
After the hand-bot grabs the book, it lets go of the shelf. Suspended by the string, the bot lowers itself gently to the ground.
Speaking the same language
For this group of distinct bots to work together, they must all speak the same language. All bots have LEDs on their exteriors that allow them to flash different colours at each other to communicate. Also, an infrared signal helps each bot locate the others.
"It's like talking," Dorigo says. Just as humans localise different voices in a crowded hall, a robot can pick up an infrared beam and figure out the distance and direction to its source.
This constant communication could allow the robot swarm to adjust its actions on the fly, compensating for broken bots by reassigning tasks throughout the team.
For now, the Swarmanoid can only perform one task, as researchers focused their first experiment on group communication and hand-bot transportation. But Dorigo envisions the swarm as the silent safety crew of the future.
If there's a fire in a building, some bots can fly around looking for injured people, while others extinguish the flames. "Obviously the mechanics, capacity and control of the robots need to be much improved before they can be applied in the real world," Dorigo says.
Swarmanoid navigation: spatio-temporal organization through reinforcement learning
Uploaded by fducatelle on Nov 26, 2009
The Swarmanoid robots perform a navigation task. The eyebots in the air give directions to the footbots on the ground, which go back and forth between a source and target location. This is also referred to as a transportation problem. Footbots measure the time it takes them to travel between source and target, and report this to the eyebots. The eyebots use this feedback to update the direction they are sending the footbots in. Like this, the footbots practically solve a reinforcement learning problem together using the eyebots as stigmergic markers in the environment. The system as a whole is able to organize to follow different paths to go and come back.
Integrated swarmanoid demonstrator, part 1: Integrated Eyebot behavior. Eyebots spread out to find a target position, and form a grid on the ceiling connecting nest and target. Then they communicate directions towards the target position down to the Footbots. In this test a single Footbot follows the Eyebot instructions back and forth between nest and target. To improve the visibility of the Footbot, a tall cylinder is drawn above it (so you can follow it also when it goes behind the far walls).
Swarmanoid robots find shortest path over double bridge
Uploaded by fducatelle on Oct 22, 2009
Here our Swarmanoid adaptive navigation system is able to find the shortest path in a double bridge experiment designed to resemble that of Deneubourg et Al. in their experiments with ants.
Swarmanoid robots find efficient path through adaptive behavior.
Uploaded by fducatelle on Oct 22, 2009
Eyebots guide Footbots between a source and a target location. They maintain two policies for guiding the Footbots, one pointing towards the target and one pointing towards the source. The Eyebots sample from these policies to give instructions to the Footbots, and they observe Footbot behavior to update their policies. Specifically, if an Eyebot sees a Footbot that is travelling from the source to the target, it increases the policy pointing to the source for the direction that the Footbot is coming from. Also, it decreases the policy pointing to the target for this same direction. When an Eyebot sees Footbots performing obstacle avoidance behavior, they decrease both policies in the direction of the location of these Footbots, as obstacle avoidance behavior points to the presence of obstacles or to congestion between Footbots.
In the specific experiment we carry out here, we can see how the system learns to send the Footbots around some obstacles on the ground.
Footbots go back and forth between two locations, guided by the Eyebots overhead and by local communication with other Footbots using their range and bearing system. Lower level behaviors include obstacle avoidance, obstacle circumnavigation, and moving away from other Footbots.
Swarmanoid robot navigation in an open space scenario
Uploaded by fducatelle on Aug 21, 2009
Footbots go back and forth between two locations, guided by the Eyebots overhead and by local communication with other Footbots using their range and bearing system. Lower level behaviors include obstacle avoidance, obstacle circumnavigation, and moving away from other Footbots.
This movie contains fragments of the most important robots I have seen in 2011 and contains the following robots: Adapt, Angels, DARwin-OP, ECCE robot, F1 simulator, HANDLE Hand Biomorphic Arm, iCub, Knexo, Kuka/DLR leightweight arm, Kuka Youbot, Lampetra, Metal Sphere, Moway, Nabazmob Orchestra, Nao, Octopus, Paro, Philips Homerun, PR2, Probo, Quadrover, Reactable Life, Salamander, Scratchbot, Swarmanoid, Yaskawa, Yume Robo,...
President Obama and the business community have been at odds for months. But in July the chairman and cofounder of the Blackstone Group, one of the world’s largest private-equity firms, amped up the rhetoric. Stephen Schwarzman—the leading John McCain supporter in a firm that, in 2008, gave more money to Obama—was addressing board members of a nonprofit organization when he let loose.
Aug 15, 2010 4:00 AM EDT
Stephen A. Schwarzman, CEO and co-founder of the Blackstone Group., SHANNON STAPLETON
President Obama and the business community have been at odds for months. But in July the chairman and cofounder of the Blackstone Group, one of the world’s largest private-equity firms, amped up the rhetoric. Stephen Schwarzman—the leading John McCain supporter in a firm that, in 2008, gave more money to Obama—was addressing board members of a nonprofit organization when he let loose. “It’s a war,” Schwarzman said of the struggle with the administration over increasing taxes on private-equity firms. “It’s like when Hitler invaded Poland in 1939.”
Attendees at the board meeting (who provided details on condition that they and the organization not be identified) were shocked. “War? Hitler? Poland? A little over the top for a proposal to make hedge-fund managers pay their fair share in taxes,” one attendee says about the comments. Neither Blackstone nor the White House would discuss Schwarzman’s statement, which came in the wake of strong, but less stinging, criticism this summer of the administration from the U.S. Chamber of Commerce and the Business Roundtable.
VIDEO INSIDE
Schwarzman’s original beef with Obama grew out of a 2008 campaign promise that “carried interest”—the compensation structure of private-equity-fund managers—would be taxed as ordinary income (35 percent) instead of capital gains (15 percent). Obama and many Democrats have argued that it’s unfair for people like Schwarzman, with a net worth of about $8 billion, to pay taxes at a lower rate than their secretaries and chauffeurs. More substantively, the commissions and fees that hedge-fund managers reap (20 percent of their clients’ profits) are not, strictly speaking, capital gains because the managers themselves never held the stocks.
The issue grew more complicated this year when Congress made noises about taxing the sale of private partnerships as ordinary income, not capital gains. This aroused the opposition of thousands of real-estate ventures and other businesses far afield from the hedge-fund world. The government’s aim was to prevent tax lawyers for the partnerships from using more than a dozen loopholes to keep taxes low, but it had the effect of slowing efforts to confront the carried-interest discrepancy.
Schwarzman is also angry at the president for some of his rhetoric (Obama has talked of “fat-cat bankers”) and for not having a prominent former CEO in his cabinet or inner circle—concerns shared widely throughout the business community, including among Democrats. “Steve thinks the president lacks an intuitive feeling for the role of capital markets,” says a Wall Street executive who knows Schwarzman. “Obama is from Mars and Steve is from Venus.”
Iacocca's salary of $US18 million made him the highest-paid executive in the world.
That record is now thought to be held by Stephen Schwarzman, head of the US asset management and financial services firm Blackstone Group, who in 2008 received a package worth $US702 million. The runner-up, Larry Ellison, chief executive of business software company Oracle, made $US557 million. Almost a quarter-century after flint-eyed fictional financier Gordon Gekko explained the meaning of life ("It's all about bucks, kid. The rest is conversation"), he will return to cinema screens in April in Wall Street 2: Money Never Sleeps. Somehow the timing feels just right. http://investorshub.advfn.com/boards/read_msg.aspx?message_id=59073635
News 
Market Data 
Discover 
AI
