Transcript Using Radio Frequency Identification in Agent-Based

Using Radio Frequency
Identification in AgentBased
Manufacturing Control
Systems
By: Lamar Alston
ASQ Student Member
RFID
 (RFID) is a technology for automatic
identification and localization of items,
particularly in supply chain. Unlike bar
code technology that detects the optical
signals reflected from barcode labels,
RFID uses radio waves to transmit the
information from an RFID tag placed on
the physical object to the RFID reader.
Agent Base Controlled
System Tracking Inventory
Cambridge Packing
 The Cambridge packing
(1) Montech conveyor
(2) gate,
(3) RFID reader,
(4)docking station,
(5) Fanuc M6i robot,
(6) storage area,
(7) rack storage,
(8) gantry robot,
(9) shuttle
Significant Changes in
Control Hierarchy
 The most important aspect of the agent-based
control system is, that unlike in classical
Computer Integrated Manufacturing (CIM)
systems, there is no central control element.
 The decision-making processes are distributed
over a community of units – agents – that are
responsible for the local control of particular
parts or components of the manufacturing
equipment.
RFID Data Collection
 The core of our RFID-oriented
development is a small, self-contained
Java application called RFID Manager,
which task is to collect the RFID data
from the readers, provide basic filtering
mechanisms and temporarily store the
data in its internal memory.
RFID Data mining
 The most commonly required type of filtering
technique is the duplicity read filtering– as the
tag enters the RF field of the reader, its EPC is
reported by the reader continually as long as
the tag resides in the field.
 Another useful processing module commonly
implemented is the Logical Mapping filter,
which allows one to define separated reading
zones called logical readers and associate the
information about the physical reader and
antenna that actually read an EPC with the
appropriate logical reader entity.
Rockwell’s Agent Based
Platform
 The RFID Logix Manager provides the
EPC data to the control applications
running on an industrial PLC (Rockwell
Automation ControlLogixTM) by writing
them directly into the PLC data-table
(common data memory).
Rockwell’s Agent Based
Tool
 Rockwell Automation application Aimed
particularly at the transportation of
discrete work pieces (products) among
the manufacturing cells (machines) on
the factory shop floor using a conveyorbased transportation, we have developed
the agent-based simulation and control
system called MAST – Manufacturing
Agent Simulation Tool.

--------Rockwell Automation
RFID
Agent Examples
 Work cell agent that represents a general
manufacturing cell, e.g. a drilling/milling machine,
storage area, assembly machine, docking station, etc.
Work cells play roles of source and destination
components.
 Conveyor belt used to transport work pieces between
two other components which it is connected to.
 Diverter (crossing) agent that switches work pieces
between the conveyor belts in order to navigate them
to desired destination work cell. Each diverter agent
holds an up-to-date routing table containing the
knowledge.
Conclusion
 Agent Based Control and computer
architecture is improving quality and
saying millions by allowing visualization
into the manufacturing enterprise. This
near real time availability of information
aids a company’s decision makers to
make informed decisions