Page 24 - Vol.10

Tech
Notes
技術專文

movement in case of the existence
Figure 1. The environment grid map and expansion cells
of overlapping unexplored cell and
lost contact with adjacent robot.
Moving the robot to a target cell
Robot A without other robot’s support will
cease the network communications.
Implement a swa r m behav ior
topology controller to plan the
target cell to measure the aerosol
Robot
concentration Figure 2.
Local
Communication

Explored Cell

Occupied Cell
Design of the
Un-explored Cell
Particle Detecting
Un-explored
Expansion Cell
Sensor

Expansion Cells of Robot A
ccompany ing w ith the
progress of minimizing
Athe wire width results in
a demand for fast response system
that prov ides in for mation on
to control cleanroom contamination. particle contamination and localize
particle diameter and concentration
the source of particle contamination
The particle contamination is far of monitor system for the particle
in a vast area of cleanroom? X. Cui
different from the AMC, for the contamination control. But current
et al., set up a multiple detecting
AMC is sorted by the material’s robot system , using moving robot particle monitoring methods are
[1]
compound property. But even the not sensitive enough to measure a
equipped with contaminant sensor semiconductor cleanroom ambient,
same compound, particle form
to perform extend area exploration working offline and require extra
of contaminant may show up in task. The moving robots formed an
different diameter and thus results personnel to collect and analyze the
ad-hoc network, by data exchanging particle diameter distribution and
in different behavior. The great between the robots, the system
variability in cleanroom ambient concentration records in periods
could dynamically respond the of time, which makes difficulties
aerosol characteristics results in a
concentration gradient and direct to response on the moment of the
demand for fast response system. the source of contaminant.
We cannot use the traditional way particle contamination.
to detect contaminant dispersing in The swarm-based FLC algorithm is A perfect particle detecting sensor
the surroundings by dog tracking, deployed to search for the source of should be inex pensive, sma l l
seeking the odor sources in the hazardous gas in X. Cui ‘s research. and robust providing particle
cleanroom. But the contaminants Every single robot is equipped with contamination warning. These
a l w a ys h a v e t h e n a t u r e of a sensor and plays as a node in the concepts may come true in the
diffusion and will spread up in the ad-hoc network system Figure 1. immediate future. For this reason,
environment, thus leave the clue of They assume the sensor is always it’s an urgent need to have real
concentration gradient to localize influenced by the random noise, thus time data of characterization
the source of contaminant. We can not offer absolutely correct data. of pa r t icle conta m i nat ion i n
have been using Scanning Mobility Each robot could communicate with semiconductor cleanroom ambient.
Particle Sizer(SMPS) on particle others, but in a limited range. The In the recent past, Bukowiecki
detection for quite a long period area of the surveying environment et al, 2002, developed a real time
of time before minimized particle is adapted to orthogonal gridded characterization system of ultra fine
[2]
sensors were invented. Is it possible map, and the swarm behavior particle on highways .
to drive a SMPS to explore the is used here to control robot’s Thus we propose the application

24