measurements are obtained for each entry point, and statistics
including the mean, variance, standard deviation, and maximum
and minimum values may be determined. This analysis tests the
reliability of the RFID prototype in detecting a transmitter (tag)
that gets close to a specific plane. Fig. 2 shows a general diagram
of the major system components.
In this configuration, use of the robot enables definition of different
risky areas. Such a definition can be carried out with the RAPID
language. The setup shown in Fig. 2 would need essentially a
change in the different points pre-programmed, which is only a
software modification and does not change the experimental setup
at the hardware level. This provides important flexibility for simulating
different situations that could result in an accident, with different
kinds of machinery. The tested device is independent of the
experimental setup. The only necessary condition is that the device
must activate its output (connected to a digital input of the robot
controller) during a specific period of time. During this time, the
robot end effector (and thus the transmitter attached) is automatically
moved outside the risky area by a programmed interrupt
routine. It then waits for deactivation of the output. Thus, this time
must be long enough to allow the robot to finish moving. Under
these conditions, is possible to automate testing for different
points of the dangerous region. The circuitry holding the activation
of the output of the safety device is not complex; rather, a timing
device is sufficient. In this work, the required delay has been programmed
into the microcontroller that manages the RFID reader.
In the lab tests, each time the RFID output is activated, the position
of the end effector as generated by the ABB controller is stored.
The points to which the robot moves are pre-programmed, and are
linked to the risky region. For this work, a plastic hand and a wristband
with at least one passive RFID tag have been attached to the
extreme part of the robot. The distance between the extreme part
of the hand and the RFID tag is 180 mm, as shown in Fig. 3.