Saturday, April 30, 2011

A Great Management Revolution: Management By Robot

This article in the April 2nd Economist magazine really caught my attention. Robots are definitely doing incredible things. But, Management?

Recently, I talked with John Hanak, leader of Purdue University's tech park network. He related that in his old career, he was in the steel industry in northwest Indiana as general counsel of a major steel producer. Steel in NW Indiana had employed 30,000 union employees. Through automation in the 80's and 90's, new products were produced at higher quality, higher volume, and with 15,000 laborers. Now are we really ready to replace the Managers and Supervisors too? A scary thought.

As Schumpeter points out, the US Military is actually on the cutting edge of the implementation of robot technology with their drones and a robot that fuels itself with biomass. Nanotechnology is conceiving of truly mind blowing robots that can function on a scale barely visible to the naked eye.

Unfortunately, the lure of Schumpter's article didn't deliver. He never really discussed robots making decisions. All his examples were robots in task oriented roles. Could it be possible to program a robot to make better or more reasoned decisions than a human? Hal in the movie 2001 was fallible. But, can our research universities developing robots and computers like the recent Jeopardy robot bring technology to a level of replacing a real Manager?

Here's Schumpeter's article:


Schumpeter

I, robot-manager
Management thinkers need to ponder more about homo-robo relations

Mar 31st 2011 | from the print edition


ROBOTS have been the stuff of science fiction for so long that it is surprisingly hard to see them as the stuff of management fact. A Czech playwright, Karel Capek, gave them their name in 1920 (from the Slavonic word for “work”). An American writer, Isaac Asimov, confronted them with their most memorable dilemmas. Hollywood turned them into superheroes and supervillains. When some film critics drew up lists of Hollywood’s 50 greatest good guys and 50 greatest baddies, the only character to appear on both lists was a robot, the Terminator.

It is time for management thinkers to catch up with science-fiction writers. Robots have been doing menial jobs on production lines since the 1960s. The world already has more than 1m industrial robots. There is now an acceleration in the rates at which they are becoming both cleverer and cheaper: an explosive combination. Robots are learning to interact with the world around them. Their ability to see things is getting ever closer to that of humans, as is their capacity to ingest information and act on it. Tomorrow’s robots will increasingly take on delicate, complex tasks. And instead of being imprisoned in cages to stop them colliding with people and machines, they will be free to wander.

America’s armed forces have blazed a trail here. They now have no fewer than 12,000 robots serving in their ranks. Peter Singer, of the Brookings Institution, a think-tank, says mankind’s 5,000-year monopoly on the fighting of war is breaking down. Recent additions to the battlefield include tiny “insects” that perform reconnaissance missions and giant “dogs” to terrify foes. The Pentagon is also working on the EATR, a robot that fuels itself by eating whatever biomass it finds around it.

Related topics
The Pentagon
But the civilian world cannot be far behind. Who better to unclog sewers or suck up nuclear waste than these remarkable machines? The Japanese have made surprisingly little use of robots to clear up after the recent earthquake, given their world leadership in this area. They say that they had the wrong sort of robots in the wrong places. But they have issued a global call for robotic assistance and are likely to put more robots to work shortly.

As robots advance into the service industries they are starting to look less like machines and more like living creatures. The Paro (made by AIST, a Japanese research agency) is shaped like a baby seal and responds to attention. Honda’s robot, ASIMO, is humanoid and can walk, talk and respond to commands. Roxxxy, an American-made “sex robot”, can be programmed to appeal to all preferences, and (unlike many a real-life spouse) listens to its partner to try to improve its performance.

Until now executives have largely ignored robots, regarding them as an engineering rather than a management problem. This cannot go on: robots are becoming too powerful and ubiquitous. Companies may need to rethink their strategies as they gain access to these new sorts of workers. Do they really need to outsource production to China, for example, when they have clever machines that work ceaselessly without pay? They certainly need to rethink their human-resources policies—starting by questioning whether they should have departments devoted to purely human resources.

The first issue is how to manage the robots themselves. Asimov laid down the basic rule in 1942: no robot should harm a human. This rule has been reinforced by recent technological improvements: robots are now much more sensitive to their surroundings and can be instructed to avoid hitting people. But the Pentagon’s plans make all this a bit more complicated: many of its robots will be, in essence, killing machines.

A second question is how to manage the homo side of homo-robo relations. Workers have always worried that new technologies will take away their livelihoods, ever since the original Luddites’ fears about mechanised looms. That worry takes on a particularly intense form when the machines come with a human face: Capek’s play that gave robots their name depicted a world in which they initially brought lots of benefits but eventually led to mass unemployment and discontent. Now, the arrival of increasingly humanoid automatons in workplaces, in an era of high unemployment, is bound to provoke a reaction.

Loving the alien

So, companies will need to work hard to persuade workers that robots are productivity-enhancers, not just job-eating aliens. They need to show employees that the robot sitting alongside them can be more of a helpmate than a threat. Audi has been particularly successful in introducing industrial robots because the carmaker asked workers to identify areas where robots could improve performance and then gave those workers jobs overseeing the robots. Employers also need to explain that robots can help preserve manufacturing jobs in the rich world: one reason why Germany has lost fewer such jobs than Britain is that it has five times as many robots for every 10,000 workers.

These two principles—don’t let robots hurt or frighten people—are relatively simple. Robot scientists are tackling more complicated problems as robots become more sophisticated. They are keen to avoid hierarchies among rescue-robots (because the loss of the leader would render the rest redundant). So they are using game theory to make sure the robots can communicate with each other in egalitarian ways. They are keen to avoid duplication between robots and their human handlers. So they are producing more complicated mathematical formulae in order that robots can constantly adjust themselves to human intentions. This suggests that the world could be on the verge of a great management revolution: making robots behave like humans rather than the 20th century’s preferred option, making humans behave like robots.

No comments:

Post a Comment