|
Thursday September 11, 2008
|
Can robots be programmed and designed to follow Isaac Asimov's robot rules and live and work among people without every causing them physical harm? The question is as old as the idea of robots and people remain terrified to this day of automatons eventually becoming sentient, turning on us and doing us physical harm. This leaves aside the fact that current robot technology is likely half a century away from delivering robots that resemble CP-30. However, according to a lengthy report on Science Daily scientists at the University of Pisa's Faculty of Engineering are busing trying to develop, as a start, robot arms that can work with humans without knocking their blocks off.
This is far more complicated than it sounds. Engineers have to account for physical malfunctions, software bugs, sensor glitches and more. Even if the robot arms are functioning normally, human missteps can result in accidents--say you just happen to walk into one as its swinging around to hand a Pepsi to a co-work across the aisle. So researchers are also examining the special arm coverings and even allowing motors to decouple so that the force of the movement is not entirely transferred back to the heaviest part of the arm--which then might transfer it right back to your face. Another key area of study is proprioception, which could give the robots a human-like awareness of their surroundings and of when someone is particularly close to them.
If they successfully develop such an arm, it could lead to a full-blown, harmless robot and safer "physical human-robot interactions" (pHRI). I might handicap the plan a little better if the acronym hadn't led Pisa's Faculty of Engineering to call the whole project Phriends.
|
|
|