Automated Guided Vehicles

Chapter 5
 1. Introduction to Automated Guided
Vehicles
An Automated Guided Vehicle System (AGVS) is a
material handling system that uses independently
operated, self-propelled vehicles guided along defined
pathways in the facility floor
• Transport material from loading to unloading stations
• Highly flexible, intelligent and versatile material
handling systems.
• AGVS is a Computer-Controlled, Non-manned,
Electric Powered Vehicle Capable of Handling
Material
• A very flexible solution for the problem of integrating a
new automated transportation line into an existing
transportation environment by using automatic guided
vehicle.
• What is AGV? Driverless Vehicle, Electric motors, battery
powered, System Discipline, Programming capabilities:
- Destination
- Path selection
- Positioning
- Collision avoidance
• Types of AGV:
– Driverless trains
– Pallet trucks
– Unit load AGVs
 Driverless Automated Guided Train

• First type of AGVS to

be introduced around
1954
• Common application is
moving heavy payloads
over long distances in
warehouses and
factories without
intermediate stops
along the route
 AGV Pallet Truck
• Used to move
palletized loads along
predetermined routes
• Vehicle is backed into
loaded pallet by
worker; pallet is then
elevated from floor
• Worker drives pallet
truck to AGV guide
path and programs
destination
 Unit Load Carrier

• Used to move unit loads from station to station

• Often equipped for automatic loading/unloading of
pallets and tote pans using roller conveyors, moving
belts, or mechanized lift platforms
 2. History of AGVS
• The first AGV system was built and introduced in
1953( A modified towing tractor that was used to pull a
trailer and follow an overhead wire in a grocery
warehouse)
• AGV developed in 1954 by A.M.Barrett,Jr. Using a
overhead wire to guide a modified towing truck pulling
a trailer in a grocery warehouse. Subsequently,
commercial AGV were introduced by Barrett.
• In 1973, Volvo in Kalmar, Sweden set out to develop
non-synchronous assembly equipment as an alternative
to the conventional conveyor assembly line.  The result
was 280 computer-controlled assembly AGVs
• Volvo developed automated guided
vehicles to serve assembly platforms
for moving car bodies through its
final assembly plants.
• Later, Volvo marketed their unit
load AGVs to other car companies.
• In the 1970’s the principal guidance
technology was to induce an
electronic frequency through a wire
that was buried in the floor. ‘floor
controller’
• These first generation navigation
schemes were expensive to install.
• All floor cuts needed to follow the
exact path of the AGV.
• Introduction of a unit load vehicle
- They have the ability to serve several functions;
- a work platform,
- a transportation device, and
- a link in the control and information system
• They transport material in warehouses, factories,
mills, hospitals, and other industrial and commercial
settings
 3. Vehicle functions
• Man / vehicle functions
- Inputs made via operator panel with its keyboard and display
- Destination input to the vehicle
- Plug-in manual control and diagnosis module
• Route (destination) finding
- High vehicle intelligence
- Travel route topology stored in the vehicle
- Destination code processing
- Load-sensing and empty location recognition
• Guide track following
Guided movements using:
- optical track
- inductive track
- "free-flight" (partly guide-trackless)
- "free-navigation" (guide-trackless
• Data exchange
- Infrared
- Radio
• Special functions
- Battery reserve monitoring
- Control of battery charging
- Obstacle recognition
• Load handling
- Load acceptance
- Load depositing
- Load monitoring
- Load transfer synchronization
• Travel control
- Speed
- Safety gap maintenance
- Collision protection
 Modern AGVS
• Modern AGVs are computer
controlled vehicles with onboard
microprocessors.
• Position feedback system to
correct path
• Communication between
vehicles via system controller
– RF communication
– Electric signals
• System management computers
• Optimizing the AGV utilization
• Tracking the material in transfer
and directing the AGV traffic.
 4. Components of AGVS
• The Vehicle – No operator
• The guide path – The path for the AGV
• The control Unit – Monitors and Directs system
operations including feedback on moves,
inventory, and vehicle status.
• The computer interface – Interfaces with other
mainframe host computer, the automated storage
and retrieval system (AS/RS), and the flexible
manufacturing system.
What is in a system?
Control concept of AGVs
 5. Types of AGVSs
AGVS Towing Vehicles
• First type of AGV introduced.
• Towing vehicle is called an
automated guided tractor
• Flatbed trailers, pallet trucks,
custom trailers can be used.
• Generally, used for large
volumes (>1000 lb) and long
moving distances (>1000
feet).
Vehicle types – Tow/Tugger
AGVS Forklift Trucks
• Ability to pickup and drop
palletized load both at floor
level and on stands.
• Pickup and drop off heights
can be different
• Vehicle can position its fork
according to load stands with
different heights
• Very expensive
• Selected where complete
automation is
necessary/required.
Vehicle types - Fork
Vehicle types – Unit load
 6. AGVs Guidance system
• The goal of an AGVs guidance system keep the AGV on
track/predefined path
• One of the major advantage of AGV is ease in modification
given by the guidance system for changing the guide path at
low cost compared to conveyors, chains, etc.
• Another benefit is: guide path is flexible which means
intersection of path is possible.
• Generally, guide path does not obstruct another systems.
• The guidance systems can be selected based on the type of
AGV selected, its application, requirement and
environmental limitation
 Vehicle Guidance Technology
• Three main technologies:
– Imbedded guide wires - guide wires in the floor
emit electromagnetic signal that the vehicles
follow
– Paint strips - optical sensors on-board vehicles
track the white paint strips
– Self-guided vehicles - vehicles use a combination
of
• Dead reckoning - vehicle counts wheel turns in
given direction to move without guidance
• Beacons located throughout facility - vehicle
uses triangulation to compute locations
Vehicle Guidance Using Guide Wire
 AGV Navigation
• 1980s Non-Wire Guidance
• The introduction of laser and inertia guidance.
– Allow for increased system flexibility and
accuracy
– No need for floor alterations or production
interruption
• The principles which make it possible for an AGV to
navigate its way between any two locations are really
quite simple. All navigation methods use a path. The
vehicle is instructed to Follow a Fixed Path or Take
an Open Path
• Fixed Path Navigation Following a Path
• The paths are well marked on the floor
• The paths are continuous
• The paths are fixed, but can be changed
Creating a Path:
The principal techniques for
creating paths are to:
• Apply a narrow magnetic tape on
the surface of the floor
• Apply a narrow photo sensitive
chemical strip on the surface of
the floor
• Apply a narrow photo reflective
tape on the surface of the floor
• Bury a wire just below the
surface of the floor
• Bury a current-carrying wire just
below the surface of the floor
Steering Correction
Coils
• The vehicle steers
itself to FOLLOW
the magnetic field
surrounding the
buried wire
Path Selection
• In this illustration, a vehicle at
“A” has two choices on how to
get to “B”.  A computer either on
board the vehicle or at some
central location selects a path
based on established criteria.

• Criteria:
– The shortest distance
– The path with the least traffic
at the present time

• All of the “PATH FOLLOWING”

methods permit routing options
that include guide path switching
and merging.
 Open Path Navigation: Taking a Path

• Unlike “path following

navigation,” where the
guide paths are fixed, and
more or less permanent,
vehicles operating in the
“Take a Path” category are
actually offered more
variation if not an infinite
number of ways to navigate
the open space between
two points.
 Typical navigation systems for AGVs

• Laser triangulation
• Inertial
• Magnetic tape
• Magnetic grid
• Natural feature
• Wire
• Optical
On certain applications more than one form of
navigation may be used by a vehicle
 How do the vehicles know where they are going?

Laser triangulation navigation

systems
• Most popular method of AGV
navigation
• Reflective targets are mounted
throughout the facility at known
positions
• A laser scanner is mounted on top
of the vehicle
• The laser scanner strobes for
reflective targets
• The vehicle control algorithms
calculate the exact vehicle position
via triangulation
• Reference points are
strategically located
targets
• A beacon on top of the
vehicle emits a rotating
laser beam which is
reflected back to the
vehicle when it strikes
(sees) a target.
Inertial navigation systems

• Reference points (often magnets)

are embedded in the floor at certain
x,y coordinates in a map of the
system
• Reference points are detected by a
sensor on the vehicle as it passes
over the reference point
• A gyroscope on the vehicle
measures/maintains vehicle’s
heading
• A wheel encoder on the vehicle
calculates the distance traveled
• Vehicle uses feedback from all
three devices to determine location
• An on-board gyroscope
establishes and
maintains a vehicle’s
heading.
• Distance traveled is
calculated by an on
board encoder which
counts wheel rotations
Magnetic tape navigation
systems
• Magnetic tape is adhered
on the surface of the floor
• A sensor on underside of
vehicle detects the
magnetic tape
• Can operate off tape path
via dead reckoning
• Similar to wire guidance
(described on future slide)
Grid Navigation Systems
• Reference points (often magnets)
are embedded in the floor in a grid
pattern in the operating area
• Reference points are given x,y
coordinates which are stored in the
vehicle’s memory
• The reference points are detected
by an on-board sensing device
• A gyroscope on the vehicle
measures/maintains heading
• A wheel encoder on the vehicle
calculates the distance traveled
• Vehicle uses all three devices to
determine location
Natural feature navigation systems
• Reference images of the operating
area are recorded and stored in the
vehicle’s memory
• Uniquely identifiable, naturally
occurring features are identified in
the operating area
• Vehicle’s actual position is
calculated based on its relative
position compared to those natural
features
• A camera or laser can be used to
record features during setup and
sense features during navigation
Wire navigation systems
• Navigates using a
continuous wire embedded
in the floor
• Antennas located on the
vehicle detect signal from
the wire
• Uses encoders on wheels
to calculate distance
• Typically used in retrofits,
system replacements and
expansions
Optical navigation systems
• Chemical or tape strip is fixed
or painted to the floor
• Vehicle has an onboard sensor
which allows it to detect the path
• Some systems use an
ultraviolet (UV) light source
under vehicle to illuminate the
strip which may not be visible
with non-UV lighting
• Not typically used in plants or
warehouses because floor line
needs to be cleaned or reapplied
 Will vehicles be able to move the loads?

Master vehicles
• The most flexible vehicle type
• Able to interface with the floor
and block stacking, racking, stands
and conveyors
• Resemble typical manual lift truck
models (counterbalanced, reach,
outrigger)
• Can be fitted with typical manual
lift truck load handling attachments
(forked, clamp, single/double)
Unit load vehicles
• Very compact design with
vehicle typically directly
under the load
• Able to interface with
stands, and conveyors
• Load handling typically via
conveyor or lift on the top of
the vehicle
• May have no load handling
ability if loaded/unloaded by
external means (crane, etc)
Tow vehicles
• Sometimes called tugger
vehicle
• Tows several (typically up
to 3) wheeled carts
• Loads must be placed on
and off carts manually or via
some other automated
machinery
• Provides most economical
solution (fewer vehicles) than
solutions where only one
load is carried per trip
 7. What about Safety?
• Most industrial-use AGVs travel at a speed between 100 and
300 feet per minute
Safety elements
 7. AGVs System Design
Many issues must be considered before designing
system for an AGVS:
• Selection of guidepath and vehicle
• Guidepath layout / Flow path design
• Number of vehicles
There are several other issues regarding timing of
AGVs, dispatching rules, routes, etc. Also, there
must be interaction between design and operational
issues for system design.
 AGVs design features
Stopping Accuracy:
• Automatic load transfer- High accuracy
• Manual load transfer- Low accuracy
• Unit load transporters are used for systems requiring high
accuracy.
• Feedback system can be used to provide stopping accuracy
• Depends on the requirement/application.
Facility: Environmental compatibility, elevator, sensors, etc.
facilities must be considered while designing AGVS.
Safety features: Emergency buttons, object detection for collision
avoidance, warning signals, must be built into the system.
Maintenance:
• Preventive maintenance intervals should be specified.
• Routine and repair maintenance including lubrication, checking
systems electrical/electronic parts .
 Attributes for selection of guidance
Vehicle Related attributes:
• Cost of the vehicle system
• Cost of guidance system
• Vehicle dimensions
• Load capacity
- Maximum weight
- Maximum Volume (depending on AGV inner dimensions)
• Maximum speed at loaded/unloaded condition
• Maintenance facilities: Modular components for maintenance,
self diagnosis, etc.
• Charging related attributes such as charging time, on-line
charging.
• Turning Radius
• Position sensors
• Loading system i.e., unit load, pallet, or other attachments
required.
 Flow path design

• Type of flowpath within the layout i.e. unidirectional,

bidirectional or combination
• Type of guidepath layout
• Position of load transfer or loading /unloading stations
• Number of stoppage stations
• Storage space of the stations.
For developing a flow path design simulation software can
be used. These software takes into consideration the
layout, locations of P/D stations, timings of AGV,
material flow intensities between stations, etc.
 Number of vehicles
General notations:
• Dd = total average loaded travel distance
• De = total average empty travel distance
• Ndr = Number of deliveries required per hour
• Nd = Number of deliveries per vehicle per hour
• Th = loading and unloading time
• Tdv = total time per delivery per vehicle
• Tf = traffic factor that accounts for blocking and waiting of vehicles and at
intersections.
if only 1 vehicle than Tf =1
if Number of vehicles > 1 than Tf <1
• v = vehicle speed
Tdv = (Dd / v) + Th + (De / v)
= loaded travel time + loading/unloading time + empty travel time
Nd = ( 60*Tf ) / Tdv
Number of automated guided vehicles = Ndr / Nd
 Advantages of AGV’s
• Unobstructed movement
• Flexibility
- Locations, path, P/D points can be reprogrammed
- Easy to change guide path system
- Number of vehicles can be altered depending on requirement
• Greater reliability
- Less environmental problems
- AGV can be replaced by another, in case of failure.
• Lower investment
• Higher operating savings on long run
- Minimal labor cost
- Easy maintenance
• Easy to interface with other systems
- Best choice for AS/RS, FMS
 AGVs Applications

1. Driverless train operations - movement of large

quantities of material over long distances
2. Storage and distribution - movement of pallet loads
between shipping/receiving docks and storage racks
3. Assembly line operations - movement of car bodies and
major subassemblies (motors) through the assembly
stations
4. Flexible manufacturing systems - movement of work-
parts between machine tools
5. Miscellaneous - mail delivery and hospital supplies
 8. Analysis of Vehicle-Based Systems
• Equipment used in vehicle-based material transport
systems includes industrial trucks (both hand trucks and
powered trucks), automated guided vehicles, rail-guided
vehicles, and certain types of conveyor systems.
• These systems are commonly used to deliver individual
loads between origination and destination points.
• Two graphical tools that are useful for displaying and
analyzing data in these deliveries are the from-to chart
and the network diagram.
• The from-to chart is a table that can be used to
indicate material flow data and/or distances between
multiple locations.
• A network diagram consists of nodes and arrows, and
the arrows indicate relationships among the nodes.
 Mathematical equations
• It is assumed that the vehicle moves at a constant
velocity throughout its operation and that effects of
acceleration, deceleration' and other speed differences
are ignored.
• The time for a typical delivery cycle in the operation
of a vehicle-based transport system consists of (1)
loading at the pickup station, (2) travel time to the
drop-off station, (3) unloading at tire drop-off station,
and (4) empty travel time of the vehicle between
deliveries. The total cycle time per delivery per
vehicle is given by
where Tc : delivery cycle time, min/del;
TL : time to-load at load station, min;
Ld : distance the vehicle travels between load and unload
station, m (ft)
vc : carrier velocity, m/min (ft/min)
Tu : time to unload at unload station, min;
Le : distance the vehicle travels empty until the start of the
next delivery cycle, m (ft).
The Tc must be considered an ideal value, because it
ignores any time losses due to reliability problems, traffic
congestion, and other factors that may slow down a
delivery. In addition, not all delivery cycles are the same
• Originations and destinations may be different from one
delivery to the next, which affect the Ld and Le terms in
the equation.
• Accordingly, these terms are considered to be average
value for the loaded and empty distances traveled by the
vehicle during a shift or other period of analysis.
• The delivery cycle time Tc can be used to determine two
values of interest in a vehicle-based transport system: (1)
rate of deliveries per vehicle and (2) number of vehicles
required to satisfy a specified total delivery requirement.
• The possible time losses include (1) availability, (2)
traffic congestion, and (3) efficiency of manual drivers in
the case of manually operated trucks.
• The traffic factor Ft is defined as a parameter for estimating
the effect of these losses on system performance.
• Sources of inefficiency accounted for by the traffic factor
include waiting at intersections, blocking of vehicles (as in
an AGVS); and waiting in a queue at load/unload stations.
• If these situations do not occur, then Ft = 1.0. As blocking
increases, the value of Ft decreases. Ft is affected by the
number of vehicles in the system relative to the size of the
layout.
• If there is only one vehicle in the system, no blocking
should occur, and the traffic factor will be 1.0. For systems
with many vehicles, there will be more instances of
blocking and congestion, and the traffic factor will take a
lower value. Typical values of traffic factor for an AGVS
range between 0.85 and 1.0.
• For systems based on industrial trucks,
including both hand trucks and powered trucks
that are operated by human workers, traffic
congestion is probably not the main cause of
low operating performance.
• Instead, performance depends primarily on the
work efficiency of the operators who drive the
trucks. Worker efficiency is defined as the
actual work rate of the human operator relative
to the work rate expected under standard or
normal performance.
• Let Ew symbolize worker efficiency
• With these factors defined, the available time per hour per
vehicle can now be expressed as 60 min adjusted by A, Ft,
and Ew. That is,
AT = 60 A Ft Ew
where AT : available time, min/hr per vehicle;
A : availability; Ft: traffic factor, and
Ew : worker efficiency. The parameters A, Ft, and Ew do not
take into account poor vehicle routing, poor guide-path
layout, or poor management of the vehicles in the system.
• These factors should be minimized, but if present they are
accounted for in the values of Ld, Le, Tl, and Tu.
• Equations for the two performance parameters of interest can
now be written. The rate of deliveries per vehicle is given by
where Rdv : hourly delivery rate per vehicle, deliveries/hr
per vehicle;
• Tc: delivery cycle time computed by previous Equation,
min/del;
• and AT : the available time in t hour, adjusted for time
losses, min/hr.
• The total number of vehicles (trucks, AGVs, trolleys,
carts, etc.) needed to satisfy a specified total delivery
schedule Rf in the system can be estimated by first
calculating the total workload required and then dividing
by the available time per vehicle. Workload is defined as
the total amount of work, expressed in terms of time, that
must be accomplished by the material transport system in
t hr. This can be expressed as
WL = Rf . Tc
where WL : workload, min/hr;
• Rf : specified flow rate of total deliveries per hour
for the system, deliveries/hr;
• and Tc: delivery cycle time, min/del.
• Now the number of vehicles required to
accomplish this workload can be written as

• where nc : number of carriers (vehicles) required,

• WL : workload, min/hr; and
• AT : available time per vehicle, min/hr per vehicle.
where nc : number of carriers required,
• Rf : total delivery requirements in the system,
deliveries/hr;
• and Rdv : delivery rate per vehicle, deliveries/hr per
vehicle. Although the traffic factor accounts for delays
experienced by the vehicles, it does not include delays
encountered by a load/unload station that must wait for
the arrival of a vehicle.
• The preceding equations do not consider this idle time or
its impact on operating cost.
• Mathematical models based on queuing theory are
appropriate to analyze this more complex stochastic
situation
 Example 1
• Consider the AGVS layout in Figure . Vehicles travel
counterclockwise around the loop to deliver loads from the
load station to the unload station. Loading time at the load
station : 0.75 min, and unloading time at the unload station :
0.50 min. The following performance parameters are given:
• vehicle speed : 50 m/min, availability : 0.95, and traffic
factor : 0.90.
• Operator efficiency does not apply, so Ew : 1.0. Determine
• (a) travel distances loaded and empty,
• (b) ideal delivery cycle time, and
• (c) number of vehicles required to satisfy the delivery
demand if a total of 40 deliveries per hour must be
completed by the AGVS
• (a) Ignoring effects of slightly shorter distances around the
curves at corners of the loop, the values of Ld and Le are readily
determined from the layout to be 110 m and 80 m, respectively.
• Determining the average travel distances, Ld
and Le, requires the particular analysis of
AGVS layout and how the vehicles are
managed. For a simple loop layout such
Figure, determining these values
straightforward. For a complex AGVS layout,
the problem is more difficult. The following
example illustrates the issue.
 Example 2

The layout for this example is shown in Figure, and the from-
to chart is presented in table. The AGVS includes load station
1 where raw parts enter the system for delivery to any of three
production stations 2, 3, and 4. unload station 5 receives
finished parts from the production stations. Load and unload
times at stations 1 and 5 are each 0.5 min. Production rates for
each workstation are indicated by the delivery requirements
in Table. A complicating factor is that some parts must be
transshipped between stations 3 and 4. Vehicles move in the
direction indicated by the arrows in the figure.
Determine the average delivery distance, Ld
• Proc: processing operation, Aut: automated; Man: manual operation
• Dimensions in meter (m)
• To determine the value of Ld, a weighted average must be
calculated based on the number of trips and corresponding
distances shown in the from-to chart for the Problem:
• Determining Le the average distance a vehicle travels
empty during a delivery cycle, is more complicated. It
depends on the dispatching and scheduling methods
used to decide how a vehicle should proceed from its
last drop-off to its next pickup.
• In Figure, if each vehicle must travel back to station 1
after each drop-off at stations 2, 3, and 4, then the
empty distance between pick-ups would be very long
indeed. Le would be greater than Ld.
• On the other hand, if a vehicle could exchange a raw
work part for a finished part while stopped at a given
workstation, then empty travel time for the vehicle
would be minimized.
• However, this would require a two-position platform
at each station to enable the exchange. So this issue
must be considered in the initial design of the AGVS.
• Ideally, Le should be reduced to zero. It is highly
desirable to minimize the average distance a vehicle
travels empty through good design of the AGVS and
good scheduling of the vehicles.
• The mathematical model of vehicle-based systems
indicates that the delivery cycle time will be reduced
if Le is minimized, and this will have a beneficial
effect on the vehicle delivery rate and the number of
vehicles required to operate the system.
 EXERCISES

1. An automated guided vehicle system has an

average travel distance per delivery : 220 m and
an average empty travel distance : 160 m. Load
and unload times are each24 sec and the speed of
the AGV : 1 m/sec. Traffic factor : 0.9 and
availability : 0.94. How many vehicles are
needed to satisfy a delivery requirement of 35
deliveries/hr?
2. In Example.2, suppose that the vehicles operate according to
the following scheduling rules: (1,) vehicles delivering raw
work parts from station 1 to stations 2,3, and 4 must return
empty to station 5; and (2) vehicles picking up finished parts at
stations 2,3, and 4 for delivery to station 5 must travel empty
from station 1.
(a) Determine the empty travel distances associated with each
delivery and develop a from-to chart in the format of Table.
(b) (b) The AGVs travel at a speed of 50 m/min and the traffic
factor : 0.90. Assume reliability:100%. From Example .2,
the delivery distance Ld: 103.8m. Determine the value of
Le.
(c) (c) How many automated guided vehicles will be required
to operate the system?