| Outcome/Accomplishment:
An excavator simulator is being employed by the Georgia Institute of Technology
and North Carolina A&T State University (NCAT) for studies of system
performance under human operator control using different interface characteristics.
The simulator, developed at Georgia Tech and Purdue University, partner
institutions of the NSF-funded Engineering Research Center (ERC) for Compact
and Efficient Fluid Power (CCEFP), combines the actual cab of a Bobcat
435 mini-excavator with a full dynamic model of the excavator’s hydraulic
and mechanical systems that displays a simulated excavator arm and environment
on a 52” LCD television screen.
Impact/Benefit:
The excavator simulator models productivity, fuel efficiency, accuracy
of motion and frequency of errors as indicators of performance. In addition,
NCAT is exploring operator behavior when employing the various interfaces
to enhance compatibility between machine and operator.
Explanation/Background:
Over the years, many fluid power applications (such as excavators) have
gone through technological innovations, bringing important improvements
to machine efficiency. Despite these advancements, issues such as
high pressure, friction, containment, and constant movement continue to
present problems with controllability, leaks, and losses in efficiency.
For human operators, many systems are still manually controlled, requiring
excessive amounts of energy, intense task concentration, high skill level,
and decision-making capabilities. Complex interactions between the
operator and the system due to these requirements can lead to errors and
miscommunication. Therefore, it is important to consider the physical
and cognitive limitations of operators when designing these systems.
In order to achieve optimal overall system performance, both machine performance
and operator performance need to be improved and the effectiveness of any
design advancements needs to be investigated to better understand the human-machine
interaction. Human performance modeling provides a means to simulate
these design changes and evaluate their impact on the human operator without
developing costly prototypes. By studying human performance with
fluid power systems such as the hydraulic excavator, there is the potential
to gain insights on interactions, investigate the limitations of human
performance, and better support the needs of operators.
The fidelity of the graphics
program of the excavator simulator exceeds that of most academic simulators.
Bobcat allowed access to the CAD files of the machine so that the ERC team
could create a high-fidelity graphical model of the excavator arm.
The environment shown includes trees and bushes and the shadow of the excavator’s
arm to increase the operator’s depth perception. The graphics program
also plays a continuous engine noise soundtrack that varies in volume with
the power demand of the pumps. The excavator’s dynamics are calculated
in real time on an xPC® target. A new soil model was developed
for the simulator that accounts for any bucket trajectory through the soil.
Researchers at Purdue University made friction, line loss, and other measurements
that were included in the model and improved its fidelity.
Currently two different
input devices can be used to operate the simulation. A 6-Degrees-Of-Freedom
(DOF) commercially available haptic device, the Sensable Phantom 1.0, is
mounted on a platform that is welded to the right cabin wall. The
Phantom allows two input modalities of intuitive coordinated control, and
four different modes of force feedback have also been implemented.
The machine’s original hydraulic joysticks were removed and replaced with
two electronic joysticks donated by Sauer Danfoss. The joysticks
and phantom are mounted in a configuration that allows them to be easily
moved and relocated depending on which input method is being employed during
the simulation. |
