Cooperation of Autonomous NXT Robots Using Bluetooth Wireless Technology Page: 5

                                
Fig. 7 shows how the math block feeds the calculated
intensity reading to both the send message block, and
to the switch block containing the motors.
The receive message block is placed into a loop
where it continuously receives and    acts upon
numerical messages from the lead robot. A wire feeds
into this loop as a logic statement and serves to
facilitate the continuous receipt of messages.  A
second wire from the receive message block extends to
the switch block. A 0.25 second pause is in place in
this program as well to accommodate for the distance
between the robots. As numbers are reported to the
switch block, the follower robot is capable of making
movement based on the specific number received.
I l
Fig. 7.: Program II for following robot.
III. RESULTS AND DISCUSSION
Findings and trials with the lead and follow robots
showed  a variety  of results and performances.
Troubleshooting  and  unexpected  accommodations
became necessary for both robots in multiple areas
before a predictable and reliable program was found.
A. Line Follow Troubleshooting
Without alteration of the program or the robot
itself, the lead robot was inconsistent with its ability to
follow the black line on the floor. The arrows in Fig 8
are small and relative to the speed of the robot. When
the power to the motors is set above a certain
threshold, the turning and velocity of the robot causes
the wheels to move the sensor past the opposite edge
of the tape, thus finding the lighter color of the floor
and turning left as instructed. The larger arrows (Fig
8) represent a greater velocity and show the problem
of overrunning the width of the black tape and finding
the opposite edge before the black of the tape is
detected by the light sensor. Although exaggerated,
the arrows showing the movement and path of the
robot in each case are true to the specific scenarios

encountered. Until the power settings on both motors
were decreased significantly, the cause of erratic
turning behavior was not obvious.

Fig. 8: Left.: line follow behavior with slower motor
settings, Right.: Line follow behavior with quicker
motor settings.
Another unexpected   issue occurred with the
electrical tape itself. Often times, the light sensor
would not respond to the black tape at all. After
multiple trials with different shades, colors, materials,
and lighting scenarios, it was discovered that the light
sensor is more sensitive than originally thought to the
reflectivity of the surface it reads. The sensor was
consistent when reading the waxed surface and lighter
colored linoleum floor, but not the fresh black
electrical tape. Treatment with sand paper to dull the
electrical tape eliminated all issues regarding the light
sensor's ability to read the dark line.
B. Lead-Follow Relationship
The two robots were randomly unsuccessful at
tracing a similar path as the lead robot follows the
black line. Different complexities in the path drawn
by the electrical tape do not dictate the level of success
and margin of error between the exact paths between
the lead and follow robots. Regardless of turning
radius or the variation of turns, the margin of error and
unpredictability was all equal in magnitude.
Possible solutions to the following   problem
included the issue of battery power. It is suspected
that the lead robot's job of both, following the black
tape line and sending the Bluetooth message, may
have an effect on its performance. Also, differences in
battery charge between the lead and follow robots may
lead to their unequal velocity and/or turning radius as
the follow robot attempts to receive and act upon the
numerical messages. If a difference in charge between
the two robots was truly the case, one would hope they
still turn in the same direction.