Miller Robotic Interface II Manual de usuario descargar pdf (Pagina 8)

1 STATEMENT OF THE PROBLEM

1.1 The Need for a Telepresence Robot Interface

Robot operations are progressively becoming more important in a variety of areas, especially in

environments where humans are at risk. It is better to have a robot search a cave in

Afghanistan, patrol a building complex at night for security, or search through rubble piles for

victims of a disaster; the value of having a robot do the job, so a rescuer’s or soldier’s life is

saved, is immeasurable. “In the case of the Mexico City earthquake in 1985, 135 rescuers died;

65 of those deaths were due to rescuers searching confined spaces that flooded” [Casper, 2002].

The cases described above are examples of remote robot operations. The human operator(s)

and robot(s) are operating in different locations that are not within line of sight of each other.

In this situation, the human’s knowledge of the robot’s surroundings, location, activities and

status is gathered solely through the interface. The partnership between the human and the

robot is known as human-robot interaction (HRI). Unlike driving a remote control car where

the operator can see how the car fits into its environment, the remote robot operator has no

direct physical cues as to the robot’s state. Insufficient knowledge of the robot’s state in an

urban search and rescue (USAR) environment, for example, may result in the robot contacting a

shaky support beam which could cause a secondary collapse. Without good state awareness, the

robot can be more of a detriment to the task than a benefit.

Some researchers have tried to solve the problem of a human directing a remote robot by giving

the robot full autonomy. The DARPA Grand Challenge is a prime example of where robot

autonomy is beneficial [Thrun et al. 2006]. The off-road traversal domain allows for car sized

robots, allowing them to have many sensors, computing power and battery power than a smaller

robot. In contrast, many safety critical remote robot tasks require smaller robots as well as

human judgment and human decision making in real-time. For many applications, having a

human in the loop is a requirement, at least for the foreseeable future.

The human’s comprehension of the robot’s state and environment is often termed Situation

Awareness (SA). Endsley [1988] developed the most generally accepted definition for SA: “The

1 2 3 4 5 6 7 8 9 10 11 12 13 ... 79 80

Sin comentarios

Miller Robotic Interface II Manual de usuario Pagina 8