7/30/2019 Afar for Agv in Adfms
1/38
ALGORITHM FOR FUTUREANTICIPATIVE REASONING (AFAR) FOR
AGV IN AN AUTONOMOUS
DECENTRALIZED
FLEXIBLE MANUFACTURING SYSTEM
Abhijith GopinathS1/ M-Tech(Production Engineering)Roll No. 1
A Seminar on
7/30/2019 Afar for Agv in Adfms
2/38
CONTENTS OF PRESENTATION1. FLEXIBLE MANUFACTURING SYSTEM (FMS)
2. AUTOMATED GUIDED VEHICLE (AGV)
3. AUTOMATED GUIDED VEHICLES IN FMS
4. CENTRALISED FMS
5. AUTONOMOUS DECENTRALISED FMS (AD FMS)
6. ALGORITHM FOR FUTURE ANTICIPATIVE REASONING IN AD FMS7. MODEL OF AD FMS STUDIED
8. ALGORITHM OF HYPOTHETICAL REASONING
9. PROPOSED HYPOTHETICAL REASONING
10. ANTICIPATING AD FMS CONDITIONS
11. AFAR FOR PART INPUT
12. AGV WITH INTELLIGENT KNOWLEDGE (AGVwIK)
13. CONCLUSION
14. REFERENCE
7/30/2019 Afar for Agv in Adfms
3/38
FLEXIBLE MANUFACTURING SYSTEMS
Flexibility in manufacturing means the ability to allow
variation in parts assembly and variations in process
sequence, change the production volume and change the
design of certain product being manufactured.
7/30/2019 Afar for Agv in Adfms
4/38
FLEXIBLE MANUFACTURING SYSTEM
The primarycharacteristic of FMSis that it integratesthe following:
1. Storage
2. ManufacturingMachines
3. Inspection
4. Tooling
5. MaterialsHandlingEquipments
7/30/2019 Afar for Agv in Adfms
5/38
ADVANTAGES OF FMS
To reduce set up and queuetimes
Improve efficiency
Reduce time for productcompletion
Utilize workers better Improve product routing
Produce a variety of Itemsunder one roof
Improve product quality
Serve a variety of vendorssimultaneously
Produce more product morequickly
7/30/2019 Afar for Agv in Adfms
6/38
BASIC COMPONENTS OF FMS
1. WORKSTATIONS
In present day applications the workstations are CNC machine
tools that perform machining operation on part families.
2. COMPUTER CONTROL SYSTEM
These systems are used to coordinate the activities of the
processing stations and the material handling system in the FMS.
Multi axis CNC Laser Mini
Workstation
The interface for a computer controlled
traffic control system
7/30/2019 Afar for Agv in Adfms
7/38
BASIC COMPONENTS OF FMS
3. MATERIAL HANDLING AND STORAGE SYSTEM
Functions
1. Random, independent movement of parts between
stations.2. Handle a variety of part configurations.
3. Temporary storage.
4. Convenient access for loading and unloading.
5. Compatible with computer control.
7/30/2019 Afar for Agv in Adfms
8/38
SOME MATERIAL HANDLING EQUIPMENTS
ConveyorBelts
Forklift Trucks
Cranes for Bulk
MaterialHandling
Robotic Arms
Automated Guided
Vehicles
7/30/2019 Afar for Agv in Adfms
9/38
An Automatic Guided Vehicle (AGV) is a
mobile robot that follows markers or
wires in the floor, or uses vision or
lasers.
They are most often used in industrial
applications to move materials around a
manufacturing facility or a warehouse.
The most important advantage of AGV
is that it can be optimally integrated to
any manufacturing system.
7/30/2019 Afar for Agv in Adfms
10/38
AGV SYSTEM
1 = Automated Guided Vehicle
2 = Management system
3 = Data transmission
4 = Guide track
(laser, inductive, optical)
5 = Loading/ Unloading Points
6 = Load handling equipment
7 = Machining Centres
12
3
4
5
66
7
7/30/2019 Afar for Agv in Adfms
11/38
TASK ALLOCATION TO AGV
Current
Position
Collection
Point
Delivery
Point
The shop floor layout is fed
into the AGV as a .dft
program.
Based on the instructionreceived the AGV moves
along the specified path and
performs the operation
The AGVs are
provided with sensors
to prevent collision.
At intersections AGVs
pass on a First come
First Pass Basis.
7/30/2019 Afar for Agv in Adfms
12/38
MAIN APPLICATIONS OF AGV
Repetitive movement ofmaterials over a distance
Regular delivery of stable loads
When on-time delivery iscritical and late deliveries arecausing inefficiency
Processes where trackingmaterial is important
AGVs transporting Work in Process jobs
AGVs transporting a palette of
finished goodsAGV used to move cargo in a port
7/30/2019 Afar for Agv in Adfms
13/38
AUTOMATED GUIDED VEHICLES IN FMS
An AGV based material handling system is designed and
implemented to impart flexibility and efficiency to the productionsystem.
An effective AGV controller is needed to monitor the equipmentstatus and route the work piece movement, so that the rightmaterial can be moved to the right place at the right time.
The routing algorithms for AGV are different for the 2 types of
FMS:1. Centralised FMS
2. Autonomous Decentralised FMS
7/30/2019 Afar for Agv in Adfms
14/38
CENTRALISED FMS
The route planning of AGV systems is determined by centralised decisionmaking system, which controls the entire shopfloor.
The routing instructions are sent to different AGVs to perform the task.
Central Computer
AGV
Onboard Controller Guide Path
Position
7/30/2019 Afar for Agv in Adfms
15/38
AUTONOMOUS DECENTRALISED FMS
The basic components of ADFMS are listed below:
1. Several CNC Machine Tools
2. Robots,
3. Transportation systems(AGVs)
4. Computer systems
5. Controllers and
6. Warehouses.
Each of these componentscommunicates andexchanges their informationwhile they decide on what
action to perform next.
Controller
Controller
Controller
Controller
Computer
Computer
Computer
CNC M/C
Robot
CNC M/C
AGV
LAN
7/30/2019 Afar for Agv in Adfms
16/38
ALGORITHM FOR FUTURE ANTICIPATIVE REASONING
IN AUTONOMOUS DECENTRALISED FMS
In the autonomous decentralized system, the AGV routing is
generated by several decision making subsystems.
The original problem is decomposed into an individual
routing problem for each AGV.
In this paper, a technique for anticipating the next action of
AGV is proposed.
This includes an advance prediction of action in a few steps,
which will be able to enhance the efficiency condition of the
overall FMS.
In this way, we develop an Algorithm for Future Anticipative
Reasoning (AFAR) of the next action decision of AGV.
7/30/2019 Afar for Agv in Adfms
17/38
MODEL OF AD FMS STUDIEDThe AD-FMS consists of multiple
agents inside a factory that is shown
in figure.The agent can be divided into a parts
warehouse, product warehouse,
transportation systems for material
handling (AGV) and several MCs.
The movement of the AGV inside the
FMS is restricted on the dashed linegrid with equal speeds.
The MCs can machine several types
of parts
The machining time for each type of
machining process is fixed.
Moreover, there also exist multiple
types of MC that can perform the
same machining task. The similar
types of MC are represented as MC1,
MC2, etc.
Each MC in these groups is identified
by MC1-1, MC1-2, etc.
7/30/2019 Afar for Agv in Adfms
18/38
MODEL OF AD FMS STUDIED
The information exchange and cooperation between each agent
in this AD FMS is described as follows:
The AGV transmits the information of what type of part and
where it is going,
the parts warehouse transmits the information of what type ofpart that it prepares,
the MC transmits the information of what type of part that is
currently machined and the time remaining to finish the
machining process.
All the information that is transmitted is taken by the needed
agent as materials to perform the next action.
7/30/2019 Afar for Agv in Adfms
19/38
ALGORITHM OF HYPOTHETICAL REASONING
What is Hypothetical Reasoning?
Considering different assumptions in order to see what
follows from them. In other words, reasoning about
alternative possible worlds (i.e., states of the world),
regardless of their resemblance to the actual world.
The AFARs real-time production scheduling is done by using
2 types of hypothetical reasoning:
1. Action Decision Hypothetical Reasoning (ADHR) that
decides where the AGV will move to.
2. Part Input Hypothetical Reasoning (PIHR) that decides
the kinds of parts to be input onto the production floor.
7/30/2019 Afar for Agv in Adfms
20/38
OPERATION CONDITIONS OF AGV
The operating condition of AGV is eternally broadening like a tree
structure, where the node is assumed as the next AGV action.
7/30/2019 Afar for Agv in Adfms
21/38
PROPOSED HYPOTHETICAL REASONING PROCESS Step1: The existing AGV hypothesis
depth is set as 0. Step2: For the next hypothesis
depth, if the selected branch
is FALSE, then the farthest
left side branch is selected
and assumed to be TRUTH. Step3: Run simulation to the
selected branch.
Step4: Based on the simulation
result, the selection branch is
judged whether it is TRUTHor not.
Step5: If the simulation result is
FALSE, then go to STEP 6. If it
is TRUTH, then go to STEP 8.
7/30/2019 Afar for Agv in Adfms
22/38
PROPOSED HYPOTHETICAL REASONING PROCESS
Step6: The branch to the rightis then selected and
assumed to be TRUE,
and go to Step 3.
Step7: Go up to another
depth of hypothesis. Step8: If the hypothesis is a
set value then go to
Step 9, if not go to
Step 2.
Step9: If the selection branch
becomes TRUE, then
STOP.
7/30/2019 Afar for Agv in Adfms
23/38
ANTICIPATING THE AD-FMS CONDITIONS
We propose a novel idea of forecasting future conditions in the
AD-FMS. The terms that define AFAR are given below:
1. Standard of TRUTH/FALSE judgments.
2. Function of machine selection priority [M(MCN)].
3. Function of parts warehouse selection priority [Fp ].
4. Function of product warehouse selection priority [Ff].
5. Value of part selection priority [V(n)].6. Value of Task Decentralisation [Fd (MCN)].
7/30/2019 Afar for Agv in Adfms
24/38
STANDARD OF TRUTH/FALSE JUDGMENTS
From the result of IHS that shows the operation rates of the
AD-FMS, we can judge whether the hypothetical reasoning
has any contradictions or not based on the following 6
standards.
If the standard is not achieved, then it is judged that a
contradiction has occurred.
[Std 1]: Total MC operation rates are above 75%
[Std 2]: Total MC operation rates are above 50%
[Std 3]: Total MC operation rates are above 25%
[Std 4]: Total MC operation rates are above 0%
[Std 5]: Total AGV operation rates are above 50%
[Std 6]: Total AGV operation rates are above 0%
7/30/2019 Afar for Agv in Adfms
25/38
FUNCTION OF MACHINE SELECTION PRIORITY
The MC with higher selection priority means that it still has
many tasks remaining and is given a top priority to be selected
as TRUTH.
7/30/2019 Afar for Agv in Adfms
26/38
FUNCTION OF PART WAREHOUSE SELECTION PRIORITY
Where,
max.parts.N : Maximum number of possible part input
AllProPt : Numbers of all parts at AGV or MC
destination.N : Total numbers of destinations, i.e., AGV, MC,
parts and product warehouse.
It is a value of determining the number of parts that are under
machining or transferring process,
7/30/2019 Afar for Agv in Adfms
27/38
FUNCTION OF PRODUCT WAREHOUSE SELECTION
PRIORITY
Ff= destination.N Fp
Where,
destination.N : Total numbers of destinations, i.e., AGV, MC,
parts and product warehouse.
Fp : Function of Part Warehouse Selection Priority.
7/30/2019 Afar for Agv in Adfms
28/38
VALUE OF PART SELECTION PRIORITY
Where,
TotalProT(n) : Total time for processing part n
ProRate(n) : Production rates of product n
CompPt(n) : Numbers of product n that have completed
machining process
AllCompPt : Numbers of all that have completed
machining process
ProPt(n) : Numbers of part n at AGV or MC
AllProPt : Numbers of all parts at AGV or MC
7/30/2019 Afar for Agv in Adfms
29/38
VALUE OF TASK DECENTRALIZATION
where,
MC.Efficiency : Operation rates of MCN(%)
A MC with a lower production rate has a higher value of task decentralization,
that results in the MC to be selected easier.
7/30/2019 Afar for Agv in Adfms
30/38
ALGORITHM FOR FUTURE ANTICIPATIVE
RESPONSE (AFAR) FOR PART INPUT
Algorithm for parts input and AGV Decision:Step 1: Add one depth to the hypothesis depth.
Step 2: Calculate part selection order V(n).
Step 3: Select the branch with the highest value for V(n).
Step 4: Calculate M(MCN), Fp , Ffand Fd for the selected branch.Step5: The node with the maximum values of M(MCN) is
regarded as TRUE. AGV has to move towards that node,
collect the part and take it to the machine having
highest value of Fd. Add one depth to the hypothesis depth.
Step6: If the value of M(MCN) is less than zero then select the
branch with the next highest value of V(n) and go to Step 4. If
all branches are completed go to Step 7.
Step 7: Stop.
7/30/2019 Afar for Agv in Adfms
31/38
AGV WITH INTELLIGENT KNOWLEDGE (AGV-wIK)
In this paper, AGVs are considered as intelligent agents that are
able to adopt knowledge, transmit their information to each otherand understand mutually.
If one AGV can understand the behavior of another AGV, it is
possible to avoid collision, and to cooperate in their tasks.
Here, we define each AGV as having 6 types of intelligent
knowledge:
1. Routing knowledge
2. Self knowledge
3. Others knowledge
4. Answer knowledge
5. Avoidance knowledge.
6. Sending knowledge
Long Term Memory
Short Term Memory
7/30/2019 Afar for Agv in Adfms
32/38
APPLICATION OF AFAR IN OUR NEIGHBOURHOOD
This concept gains importance as the International ContainerTransshipment Terminal at Vallarpadam is looking forward to
automation of cargo handling. The proposed VISL is also planning to provide the same facility.
AGVs have been successfully employed for material handling in manyports worldwide.
AFAR is applicable for AGVs employed in container terminals.
The incorporation of AFAR will dramatically improve the prospects ofthese projects.
Vallarpadam ICTT
Railway line to Vallarpadam
exclusively for container movement
7/30/2019 Afar for Agv in Adfms
33/38
APPLICATION IN CONTAINER TERMINALS
Here we have ExportStorage, Gate BufferStorage and Train BufferStorage instead of PartWarehouses.
Product Warehouse isreplaced by ImportStorage.
The cargo movementsare actuated by AGVs.
Huge volumes of cargocan be handled byincorporating thistechnique.
7/30/2019 Afar for Agv in Adfms
34/38
CONCLUSION
The branch of automation is witnessing major
advancements.
In this paper, a technique was proposed to anticipate the
next action of AGV, which will improve the efficiency of
Autonomous Decentralized Flexible Manufacturing System
(AD-FMS) Environment. We have also discussed on how to apply this technique in
automated container terminals.
AFAR will be a milestone in the field of logistics.
7/30/2019 Afar for Agv in Adfms
35/38
REFERENCE
This presentation was based on a research paper published in World
Academy of Science, Engineering and Technology Journal 09/2007 issue
7/30/2019 Afar for Agv in Adfms
36/38
OTHER REFERENCES
1. www.wikipedia.com
2. M. P. Groover, Automation, Production System and
Computer Integrated Manufacturing
3. P Radhakrishnan, S Subramanian, V Raju, CAD/CAM/CIM
http://www.wikipedia.com/http://www.wikipedia.com/7/30/2019 Afar for Agv in Adfms
37/38
Questions?
7/30/2019 Afar for Agv in Adfms
38/38
Thank You